Compare commits
8 Commits
Author | SHA1 | Date |
---|---|---|
Dan Howe | 64d6ca3015 | 5 years ago |
Dan Howe | c8fb90d30f | 5 years ago |
Dan Howe | e8ba383dfd | 5 years ago |
Dan Howe | 4857456b8a | 5 years ago |
Dan Howe | d336196e24 | 5 years ago |
Dan Howe | 02511aebc2 | 5 years ago |
Dan Howe | 4f6fb13ff2 | 5 years ago |
Dan Howe | 4cb144d483 | 5 years ago |
@ -1,2 +0,0 @@
|
||||
# Ignore folder content
|
||||
tests/*
|
@ -1,20 +0,0 @@
|
||||
|
||||
engines:
|
||||
duplication:
|
||||
enabled: true
|
||||
- mass = 50
|
||||
config:
|
||||
languages:
|
||||
python:
|
||||
mass_threshold: 40
|
||||
fixme:
|
||||
enabled: true
|
||||
radon:
|
||||
enabled: true
|
||||
ratings:
|
||||
paths:
|
||||
- "**.py"
|
||||
exclude_paths:
|
||||
- wafo/tests/
|
||||
- wafo/MSPPT.py
|
||||
- wafo/MSO.py
|
@ -1,25 +0,0 @@
|
||||
# .coveragerc to control coverage.py
|
||||
[run]
|
||||
branch = True
|
||||
source = wafo
|
||||
omit =
|
||||
tests/*
|
||||
source/*
|
||||
|
||||
[report]
|
||||
# Regexes for lines to exclude from consideration
|
||||
exclude_lines =
|
||||
# Have to re-enable the standard pragma
|
||||
pragma: no cover
|
||||
|
||||
# Don't complain about missing debug-only code:
|
||||
def __repr__
|
||||
if self\.debug
|
||||
|
||||
# Don't complain if tests don't hit defensive assertion code:
|
||||
raise AssertionError
|
||||
raise NotImplementedError
|
||||
|
||||
# Don't complain if non-runnable code isn't run:
|
||||
if 0:
|
||||
if __name__ == .__main__.:
|
@ -1,42 +0,0 @@
|
||||
# Temporary and binary files
|
||||
*~
|
||||
*.py[cod]
|
||||
*.so
|
||||
*.cfg
|
||||
!setup.cfg
|
||||
*.orig
|
||||
*.log
|
||||
*.pot
|
||||
__pycache__/*
|
||||
.cache/*
|
||||
.*.swp
|
||||
*.mod
|
||||
|
||||
# Project files
|
||||
.ropeproject
|
||||
.project
|
||||
.pydevproject
|
||||
.settings
|
||||
.idea
|
||||
|
||||
# Package files
|
||||
*.egg
|
||||
*.eggs/
|
||||
.installed.cfg
|
||||
*.egg-info
|
||||
|
||||
# Unittest and coverage
|
||||
htmlcov/*
|
||||
.coverage
|
||||
.tox
|
||||
junit.xml
|
||||
coverage.xml
|
||||
|
||||
# Build and docs folder/files
|
||||
build/*
|
||||
dist/*
|
||||
sdist/*
|
||||
docs/_rst/*
|
||||
docs/_build/*
|
||||
cover/*
|
||||
MANIFEST
|
@ -1,23 +0,0 @@
|
||||
doc-warnings: yes
|
||||
test-warnings: no
|
||||
strictness: high
|
||||
|
||||
pylint:
|
||||
disable:
|
||||
- too-few-public-methods
|
||||
- no-self-use
|
||||
- too-many-instance-attributes
|
||||
- invalid-name
|
||||
- missing-docstring
|
||||
- star-args
|
||||
- logging-format-interpolation
|
||||
- bad-builtin
|
||||
ignore-paths:
|
||||
- docs
|
||||
pep8:
|
||||
disable:
|
||||
- E731
|
||||
- E126
|
||||
- E128
|
||||
- E115
|
||||
- N802
|
@ -1,400 +0,0 @@
|
||||
[MASTER]
|
||||
|
||||
# Specify a configuration file.
|
||||
#rcfile=
|
||||
|
||||
# Python code to execute, usually for sys.path manipulation such as
|
||||
# pygtk.require().
|
||||
#init-hook=
|
||||
|
||||
# Add files or directories to the blacklist. They should be base names, not
|
||||
# paths.
|
||||
ignore=CVS
|
||||
|
||||
# Add files or directories matching the regex patterns to the blacklist. The
|
||||
# regex matches against base names, not paths.
|
||||
ignore-patterns=
|
||||
|
||||
# Pickle collected data for later comparisons.
|
||||
persistent=yes
|
||||
|
||||
# List of plugins (as comma separated values of python modules names) to load,
|
||||
# usually to register additional checkers. (PAB)
|
||||
load-plugins=pylint.extensions.mccabe
|
||||
|
||||
# Maximum threshold for cyclomatic complexity (PAB)
|
||||
max-complexity=10
|
||||
|
||||
# (PAB)
|
||||
consider-iterating-dictionary=yes
|
||||
|
||||
# Use multiple processes to speed up Pylint.
|
||||
jobs=1
|
||||
|
||||
# Allow loading of arbitrary C extensions. Extensions are imported into the
|
||||
# active Python interpreter and may run arbitrary code.
|
||||
unsafe-load-any-extension=no
|
||||
|
||||
# A comma-separated list of package or module names from where C extensions may
|
||||
# be loaded. Extensions are loading into the active Python interpreter and may
|
||||
# run arbitrary code (PAB)
|
||||
extension-pkg-whitelist=numpy
|
||||
|
||||
# Allow optimization of some AST trees. This will activate a peephole AST
|
||||
# optimizer, which will apply various small optimizations. For instance, it can
|
||||
# be used to obtain the result of joining multiple strings with the addition
|
||||
# operator. Joining a lot of strings can lead to a maximum recursion error in
|
||||
# Pylint and this flag can prevent that. It has one side effect, the resulting
|
||||
# AST will be different than the one from reality. This option is deprecated
|
||||
# and it will be removed in Pylint 2.0.
|
||||
optimize-ast=no
|
||||
|
||||
|
||||
[MESSAGES CONTROL]
|
||||
|
||||
# Only show warnings with the listed confidence levels. Leave empty to show
|
||||
# all. Valid levels: HIGH, INFERENCE, INFERENCE_FAILURE, UNDEFINED
|
||||
confidence=
|
||||
|
||||
# Enable the message, report, category or checker with the given id(s). You can
|
||||
# either give multiple identifier separated by comma (,) or put this option
|
||||
# multiple time (only on the command line, not in the configuration file where
|
||||
# it should appear only once). See also the "--disable" option for examples.
|
||||
#enable=
|
||||
|
||||
# Disable the message, report, category or checker with the given id(s). You
|
||||
# can either give multiple identifiers separated by comma (,) or put this
|
||||
# option multiple times (only on the command line, not in the configuration
|
||||
# file where it should appear only once).You can also use "--disable=all" to
|
||||
# disable everything first and then reenable specific checks. For example, if
|
||||
# you want to run only the similarities checker, you can use "--disable=all
|
||||
# --enable=similarities". If you want to run only the classes checker, but have
|
||||
# no Warning level messages displayed, use"--disable=all --enable=classes
|
||||
# --disable=W"
|
||||
# disable=locally-disabled
|
||||
disable=import-star-module-level,old-octal-literal,oct-method,print-statement,unpacking-in-except,parameter-unpacking,backtick,old-raise-syntax,old-ne-operator,long-suffix,dict-view-method,dict-iter-method,metaclass-assignment,next-method-called,raising-string,indexing-exception,raw_input-builtin,long-builtin,file-builtin,execfile-builtin,coerce-builtin,cmp-builtin,buffer-builtin,basestring-builtin,apply-builtin,filter-builtin-not-iterating,using-cmp-argument,useless-suppression,range-builtin-not-iterating,suppressed-message,no-absolute-import,old-division,cmp-method,reload-builtin,zip-builtin-not-iterating,intern-builtin,unichr-builtin,reduce-builtin,standarderror-builtin,unicode-builtin,xrange-builtin,coerce-method,delslice-method,getslice-method,setslice-method,input-builtin,round-builtin,hex-method,nonzero-method,map-builtin-not-iterating,W0201
|
||||
|
||||
|
||||
[REPORTS]
|
||||
|
||||
# Set the output format. Available formats are text, parseable, colorized, msvs
|
||||
# (visual studio) and html. You can also give a reporter class, eg
|
||||
# mypackage.mymodule.MyReporterClass.
|
||||
output-format=parseable
|
||||
# output-format=text
|
||||
|
||||
# Put messages in a separate file for each module / package specified on the
|
||||
# command line instead of printing them on stdout. Reports (if any) will be
|
||||
# written in a file name "pylint_global.[txt|html]". This option is deprecated
|
||||
# and it will be removed in Pylint 2.0.
|
||||
files-output=no
|
||||
|
||||
# Tells whether to display a full report or only the messages
|
||||
reports=yes
|
||||
|
||||
# Python expression which should return a note less than 10 (10 is the highest
|
||||
# note). You have access to the variables errors warning, statement which
|
||||
# respectively contain the number of errors / warnings messages and the total
|
||||
# number of statements analyzed. This is used by the global evaluation report
|
||||
# (RP0004).
|
||||
evaluation=10.0 - ((float(5 * error + warning + refactor + convention) / statement) * 10)
|
||||
|
||||
# Template used to display messages. This is a python new-style format string
|
||||
# used to format the message information. See doc for all details
|
||||
#msg-template=
|
||||
|
||||
|
||||
[BASIC]
|
||||
|
||||
# List of builtins function names that should not be used, separated by a comma
|
||||
bad-functions=map,filter,input
|
||||
|
||||
# Good variable names which should always be accepted, separated by a comma
|
||||
good-names=i,j,k,m,n,ex, Run, _, x, y, z, u, v, w, a, b, c, f, g, h, p, q, r, s, t
|
||||
|
||||
# Bad variable names which should always be refused, separated by a comma
|
||||
bad-names=foo,bar,baz,toto,tutu,tata
|
||||
|
||||
# Colon-delimited sets of names that determine each other's naming style when
|
||||
# the name regexes allow several styles.
|
||||
name-group=
|
||||
|
||||
# Include a hint for the correct naming format with invalid-name
|
||||
#include-naming-hint=no
|
||||
include-naming-hint=yes
|
||||
|
||||
# List of decorators that produce properties, such as abc.abstractproperty. Add
|
||||
# to this list to register other decorators that produce valid properties.
|
||||
property-classes=abc.abstractproperty
|
||||
|
||||
# Regular expression matching correct function names
|
||||
function-rgx=[a-z_][a-z0-9_]{2,30}$
|
||||
|
||||
# Naming hint for function names
|
||||
function-name-hint=[a-z_][a-z0-9_]{2,30}$
|
||||
|
||||
# Regular expression matching correct variable names
|
||||
variable-rgx=[a-z_][a-z0-9_]{2,30}$
|
||||
|
||||
# Naming hint for variable names
|
||||
variable-name-hint=[a-z_][a-z0-9_]{2,30}$
|
||||
|
||||
# Regular expression matching correct dummy-variable names (PAB)
|
||||
dummy-variables-rgx=^_
|
||||
|
||||
# Regular expression matching correct constant names
|
||||
const-rgx=(([A-Z_][A-Z0-9_]*)|(__.*__))$
|
||||
|
||||
# Naming hint for constant names
|
||||
const-name-hint=(([A-Z_][A-Z0-9_]*)|(__.*__))$
|
||||
|
||||
# Regular expression matching correct attribute names
|
||||
attr-rgx=[a-z_][a-z0-9_]{2,30}$
|
||||
|
||||
# Naming hint for attribute names
|
||||
attr-name-hint=[a-z_][a-z0-9_]{2,30}$
|
||||
|
||||
# Regular expression matching correct argument names
|
||||
argument-rgx=[a-z_][a-z0-9_]{2,30}$
|
||||
|
||||
# Naming hint for argument names
|
||||
argument-name-hint=[a-z_][a-z0-9_]{2,30}$
|
||||
|
||||
# Regular expression matching correct class attribute names
|
||||
class-attribute-rgx=([A-Za-z_][A-Za-z0-9_]{2,30}|(__.*__))$
|
||||
|
||||
# Naming hint for class attribute names
|
||||
class-attribute-name-hint=([A-Za-z_][A-Za-z0-9_]{2,30}|(__.*__))$
|
||||
|
||||
# Regular expression matching correct inline iteration names
|
||||
inlinevar-rgx=[A-Za-z_][A-Za-z0-9_]*$
|
||||
|
||||
# Naming hint for inline iteration names
|
||||
inlinevar-name-hint=[A-Za-z_][A-Za-z0-9_]*$
|
||||
|
||||
# Regular expression matching correct class names
|
||||
class-rgx=[A-Z_][a-zA-Z0-9]+$
|
||||
|
||||
# Naming hint for class names
|
||||
class-name-hint=[A-Z_][a-zA-Z0-9]+$
|
||||
|
||||
# Regular expression matching correct module names
|
||||
module-rgx=(([a-z_][a-z0-9_]*)|([A-Z][a-zA-Z0-9]+))$
|
||||
|
||||
# Naming hint for module names
|
||||
module-name-hint=(([a-z_][a-z0-9_]*)|([A-Z][a-zA-Z0-9]+))$
|
||||
|
||||
# Regular expression matching correct method names
|
||||
method-rgx=[a-z_][a-z0-9_]{2,30}$
|
||||
|
||||
# Naming hint for method names
|
||||
method-name-hint=[a-z_][a-z0-9_]{2,30}$
|
||||
|
||||
# Regular expression which should only match function or class names that do
|
||||
# not require a docstring.
|
||||
no-docstring-rgx=^_
|
||||
|
||||
# Minimum line length for functions/classes that require docstrings, shorter
|
||||
# ones are exempt.
|
||||
docstring-min-length=-1
|
||||
|
||||
|
||||
[ELIF]
|
||||
|
||||
# Maximum number of nested blocks for function / method body
|
||||
max-nested-blocks=5
|
||||
|
||||
|
||||
[FORMAT]
|
||||
|
||||
# Maximum number of characters on a single line.
|
||||
max-line-length=100
|
||||
|
||||
# Regexp for a line that is allowed to be longer than the limit.
|
||||
ignore-long-lines=^\s*(# )?<?https?://\S+>?$
|
||||
|
||||
# Allow the body of an if to be on the same line as the test if there is no
|
||||
# else.
|
||||
single-line-if-stmt=no
|
||||
|
||||
# List of optional constructs for which whitespace checking is disabled. `dict-
|
||||
# separator` is used to allow tabulation in dicts, etc.: {1 : 1,\n222: 2}.
|
||||
# `trailing-comma` allows a space between comma and closing bracket: (a, ).
|
||||
# `empty-line` allows space-only lines.
|
||||
no-space-check=trailing-comma,dict-separator
|
||||
|
||||
# Maximum number of lines in a module
|
||||
max-module-lines=1000
|
||||
|
||||
# String used as indentation unit. This is usually " " (4 spaces) or "\t" (1
|
||||
# tab).
|
||||
indent-string=' '
|
||||
|
||||
# Number of spaces of indent required inside a hanging or continued line.
|
||||
indent-after-paren=4
|
||||
|
||||
# Expected format of line ending, e.g. empty (any line ending), LF or CRLF.
|
||||
expected-line-ending-format=
|
||||
|
||||
|
||||
[LOGGING]
|
||||
|
||||
# Logging modules to check that the string format arguments are in logging
|
||||
# function parameter format
|
||||
logging-modules=logging
|
||||
|
||||
|
||||
[MISCELLANEOUS]
|
||||
|
||||
# List of note tags to take in consideration, separated by a comma.
|
||||
notes=FIXME,XXX,TODO
|
||||
|
||||
|
||||
[SIMILARITIES]
|
||||
|
||||
# Minimum lines number of a similarity.
|
||||
min-similarity-lines=4
|
||||
|
||||
# Ignore comments when computing similarities.
|
||||
ignore-comments=yes
|
||||
|
||||
# Ignore docstrings when computing similarities.
|
||||
ignore-docstrings=yes
|
||||
|
||||
# Ignore imports when computing similarities.
|
||||
ignore-imports=no
|
||||
|
||||
|
||||
[SPELLING]
|
||||
|
||||
# Spelling dictionary name. Available dictionaries: none. To make it working
|
||||
# install python-enchant package.
|
||||
spelling-dict=
|
||||
|
||||
# List of comma separated words that should not be checked.
|
||||
spelling-ignore-words=
|
||||
|
||||
# A path to a file that contains private dictionary; one word per line.
|
||||
spelling-private-dict-file=
|
||||
|
||||
# Tells whether to store unknown words to indicated private dictionary in
|
||||
# --spelling-private-dict-file option instead of raising a message.
|
||||
spelling-store-unknown-words=no
|
||||
|
||||
|
||||
[TYPECHECK]
|
||||
|
||||
# Tells whether missing members accessed in mixin class should be ignored. A
|
||||
# mixin class is detected if its name ends with "mixin" (case insensitive).
|
||||
ignore-mixin-members=yes
|
||||
|
||||
# List of module names for which member attributes should not be checked
|
||||
# (useful for modules/projects where namespaces are manipulated during runtime
|
||||
# and thus existing member attributes cannot be deduced by static analysis. It
|
||||
# supports qualified module names, as well as Unix pattern matching.
|
||||
ignored-modules=
|
||||
|
||||
# List of class names for which member attributes should not be checked (useful
|
||||
# for classes with dynamically set attributes). This supports the use of
|
||||
# qualified names.
|
||||
ignored-classes=optparse.Values,thread._local,_thread._local
|
||||
|
||||
# List of members which are set dynamically and missed by pylint inference
|
||||
# system, and so shouldn't trigger E1101 when accessed. Python regular
|
||||
# expressions are accepted.
|
||||
generated-members=
|
||||
|
||||
# List of decorators that produce context managers, such as
|
||||
# contextlib.contextmanager. Add to this list to register other decorators that
|
||||
# produce valid context managers.
|
||||
contextmanager-decorators=contextlib.contextmanager
|
||||
|
||||
|
||||
[CLASSES]
|
||||
|
||||
# List of method names used to declare (i.e. assign) instance attributes.
|
||||
defining-attr-methods=__init__,__new__,setUp
|
||||
|
||||
# List of valid names for the first argument in a class method.
|
||||
valid-classmethod-first-arg=cls
|
||||
|
||||
# List of valid names for the first argument in a metaclass class method.
|
||||
valid-metaclass-classmethod-first-arg=mcs
|
||||
|
||||
# List of member names, which should be excluded from the protected access
|
||||
# warning.
|
||||
exclude-protected=_asdict,_fields,_replace,_source,_make
|
||||
|
||||
|
||||
[DESIGN]
|
||||
|
||||
# Maximum number of arguments for function / method
|
||||
max-args=5
|
||||
|
||||
# Argument names that match this expression will be ignored. Default to name
|
||||
# with leading underscore
|
||||
ignored-argument-names=_.*
|
||||
|
||||
# Maximum number of locals for function / method body
|
||||
max-locals=15
|
||||
|
||||
# Maximum number of return / yield for function / method body
|
||||
max-returns=6
|
||||
|
||||
# Maximum number of branch for function / method body
|
||||
max-branches=12
|
||||
|
||||
# Maximum number of statements in function / method body
|
||||
max-statements=50
|
||||
|
||||
# Maximum number of parents for a class (see R0901).
|
||||
max-parents=7
|
||||
|
||||
# Maximum number of attributes for a class (see R0902).
|
||||
max-attributes=7
|
||||
|
||||
# Minimum number of public methods for a class (see R0903).
|
||||
min-public-methods=2
|
||||
|
||||
# Maximum number of public methods for a class (see R0904).
|
||||
max-public-methods=20
|
||||
|
||||
# Maximum number of boolean expressions in a if statement
|
||||
max-bool-expr=5
|
||||
|
||||
|
||||
[IMPORTS]
|
||||
|
||||
# Deprecated modules which should not be used, separated by a comma
|
||||
deprecated-modules=regsub,TERMIOS,Bastion,rexec
|
||||
|
||||
# Create a graph of every (i.e. internal and external) dependencies in the
|
||||
# given file (report RP0402 must not be disabled)
|
||||
import-graph=
|
||||
|
||||
# Create a graph of external dependencies in the given file (report RP0402 must
|
||||
# not be disabled)
|
||||
ext-import-graph=
|
||||
|
||||
# Create a graph of internal dependencies in the given file (report RP0402 must
|
||||
# not be disabled)
|
||||
int-import-graph=
|
||||
|
||||
# Force import order to recognize a module as part of the standard
|
||||
# compatibility libraries.
|
||||
known-standard-library=
|
||||
|
||||
# Force import order to recognize a module as part of a third party library.
|
||||
known-third-party=enchant
|
||||
|
||||
# Analyse import fallback blocks. This can be used to support both Python 2 and
|
||||
# 3 compatible code, which means that the block might have code that exists
|
||||
# only in one or another interpreter, leading to false positives when analysed.
|
||||
analyse-fallback-blocks=no
|
||||
|
||||
|
||||
[EXCEPTIONS]
|
||||
|
||||
# Exceptions that will emit a warning when being caught. Defaults to
|
||||
# "Exception"
|
||||
overgeneral-exceptions=Exception
|
@ -1,61 +0,0 @@
|
||||
# Travis configuration file using the build matrix feature
|
||||
# Read more under http://docs.travis-ci.com/user/build-configuration/
|
||||
# THIS SCRIPT IS SUPPOSED TO BE AN EXAMPLE. MODIFY IT ACCORDING TO YOUR NEEDS!
|
||||
sudo: true
|
||||
language: python
|
||||
python:
|
||||
- 2.7
|
||||
# - 3.4 # Drop support for python 3.4
|
||||
- 3.5
|
||||
- 3.6
|
||||
notifications:
|
||||
email: false
|
||||
|
||||
# Setup anaconda
|
||||
before_install:
|
||||
- if [[ "$TRAVIS_PYTHON_VERSION" == "2.7" ]]; then
|
||||
wget https://repo.continuum.io/miniconda/Miniconda-latest-Linux-x86_64.sh -O miniconda.sh;
|
||||
else
|
||||
wget https://repo.continuum.io/miniconda/Miniconda3-latest-Linux-x86_64.sh -O miniconda.sh;
|
||||
fi
|
||||
- chmod +x miniconda.sh
|
||||
- ./miniconda.sh -b
|
||||
- if [[ "$TRAVIS_PYTHON_VERSION" == "2.7" ]]; then
|
||||
export PATH=/home/travis/miniconda2/bin:$PATH;
|
||||
else
|
||||
export PATH=/home/travis/miniconda3/bin:$PATH;
|
||||
fi
|
||||
- conda update --yes conda
|
||||
- conda install --yes conda-build
|
||||
- conda create --yes -n condaenv python=$TRAVIS_PYTHON_VERSION
|
||||
- conda install --yes -n condaenv pip
|
||||
- conda config --add channels https://conda.anaconda.org/omnia
|
||||
- conda config --add channels https://conda.anaconda.org/pbrod
|
||||
- source activate condaenv
|
||||
- sudo apt-get update
|
||||
- sudo apt-get install gfortran
|
||||
# Install packages
|
||||
install:
|
||||
- conda install --yes python=$TRAVIS_PYTHON_VERSION numpy scipy nose pytest matplotlib pytest-pep8 pytest-cov mpmath pillow numba statsmodels
|
||||
- pip install funcsigs
|
||||
- pip install numdifftools
|
||||
- pip install coveralls
|
||||
- pip install codecov
|
||||
- python setup.py build
|
||||
- python setup.py install
|
||||
# - conda build conda_recipe
|
||||
# - conda install wafo --use-local
|
||||
before_script:
|
||||
- "export DISPLAY=:99.0"
|
||||
- "sh -e /etc/init.d/xvfb start"
|
||||
- sleep 3 # give xvfb some time to start
|
||||
- git config --global user.email "per.andreas.brodtkorb@gmail.com"
|
||||
- git config --global user.name "pbrod"
|
||||
script:
|
||||
- coverage run --source=wafo setup.py test
|
||||
# - py.test wafo --cov --pep8 --doctest-modules
|
||||
after_success:
|
||||
- coveralls
|
||||
- codecov
|
||||
cache:
|
||||
- apt
|
@ -1,8 +0,0 @@
|
||||
==========
|
||||
Developers
|
||||
==========
|
||||
|
||||
* Per A. Brodtkorb
|
||||
* Georg Lindgren
|
||||
* David Verelst
|
||||
|
@ -1,11 +0,0 @@
|
||||
=========
|
||||
Changelog
|
||||
=========
|
||||
|
||||
Version 0.1
|
||||
===========
|
||||
|
||||
- Feature A added
|
||||
- FIX: nasty bug #1729 fixed
|
||||
- add your changes here!
|
||||
|
@ -1,131 +0,0 @@
|
||||
environment:
|
||||
global:
|
||||
# SDK v7.0 MSVC Express 2008's SetEnv.cmd script will fail if the
|
||||
# /E:ON and /V:ON options are not enabled in the batch script intepreter
|
||||
# See: http://stackoverflow.com/a/13751649/163740
|
||||
CMD_IN_ENV: "cmd /E:ON /V:ON /C .\\appveyor\\run_with_env.cmd"
|
||||
|
||||
matrix:
|
||||
|
||||
# Python 2.7.10 is the latest version and is not pre-installed.
|
||||
|
||||
- PYTHON: "C:\\Python27.10"
|
||||
PYTHON_VERSION: "2.7.10"
|
||||
PYTHON_ARCH: "32"
|
||||
|
||||
- PYTHON: "C:\\Python27.10-x64"
|
||||
PYTHON_VERSION: "2.7.10"
|
||||
PYTHON_ARCH: "64"
|
||||
|
||||
# Pre-installed Python versions, which Appveyor may upgrade to
|
||||
# a later point release.
|
||||
# See: http://www.appveyor.com/docs/installed-software#python
|
||||
|
||||
- PYTHON: "C:\\Python27"
|
||||
PYTHON_VERSION: "2.7.x" # currently 2.7.9
|
||||
PYTHON_ARCH: "32"
|
||||
|
||||
- PYTHON: "C:\\Python27-x64"
|
||||
PYTHON_VERSION: "2.7.x" # currently 2.7.9
|
||||
PYTHON_ARCH: "64"
|
||||
|
||||
- PYTHON: "C:\\Python33"
|
||||
PYTHON_VERSION: "3.3.x" # currently 3.3.5
|
||||
PYTHON_ARCH: "32"
|
||||
|
||||
- PYTHON: "C:\\Python33-x64"
|
||||
PYTHON_VERSION: "3.3.x" # currently 3.3.5
|
||||
PYTHON_ARCH: "64"
|
||||
|
||||
- PYTHON: "C:\\Python34"
|
||||
PYTHON_VERSION: "3.4.x" # currently 3.4.3
|
||||
PYTHON_ARCH: "32"
|
||||
|
||||
- PYTHON: "C:\\Python34-x64"
|
||||
PYTHON_VERSION: "3.4.x" # currently 3.4.3
|
||||
PYTHON_ARCH: "64"
|
||||
|
||||
# Python versions not pre-installed
|
||||
|
||||
# Python 2.6.6 is the latest Python 2.6 with a Windows installer
|
||||
# See: https://github.com/ogrisel/python-appveyor-demo/issues/10
|
||||
|
||||
- PYTHON: "C:\\Python266"
|
||||
PYTHON_VERSION: "2.6.6"
|
||||
PYTHON_ARCH: "32"
|
||||
|
||||
- PYTHON: "C:\\Python266-x64"
|
||||
PYTHON_VERSION: "2.6.6"
|
||||
PYTHON_ARCH: "64"
|
||||
|
||||
- PYTHON: "C:\\Python35"
|
||||
PYTHON_VERSION: "3.5.0"
|
||||
PYTHON_ARCH: "32"
|
||||
|
||||
- PYTHON: "C:\\Python35-x64"
|
||||
PYTHON_VERSION: "3.5.0"
|
||||
PYTHON_ARCH: "64"
|
||||
|
||||
# Major and minor releases (i.e x.0.0 and x.y.0) prior to 3.3.0 use
|
||||
# a different naming scheme.
|
||||
|
||||
- PYTHON: "C:\\Python270"
|
||||
PYTHON_VERSION: "2.7.0"
|
||||
PYTHON_ARCH: "32"
|
||||
|
||||
- PYTHON: "C:\\Python270-x64"
|
||||
PYTHON_VERSION: "2.7.0"
|
||||
PYTHON_ARCH: "64"
|
||||
|
||||
install:
|
||||
- ECHO "Filesystem root:"
|
||||
- ps: "ls \"C:/\""
|
||||
|
||||
- ECHO "Installed SDKs:"
|
||||
- ps: "ls \"C:/Program Files/Microsoft SDKs/Windows\""
|
||||
|
||||
# Install Python (from the official .msi of http://python.org) and pip when
|
||||
# not already installed.
|
||||
- ps: if (-not(Test-Path($env:PYTHON))) { & appveyor\install.ps1 }
|
||||
|
||||
# Prepend newly installed Python to the PATH of this build (this cannot be
|
||||
# done from inside the powershell script as it would require to restart
|
||||
# the parent CMD process).
|
||||
- "SET PATH=%PYTHON%;%PYTHON%\\Scripts;%PATH%"
|
||||
|
||||
# Check that we have the expected version and architecture for Python
|
||||
- "python --version"
|
||||
- "python -c \"import struct; print(struct.calcsize('P') * 8)\""
|
||||
|
||||
# Upgrade to the latest version of pip to avoid it displaying warnings
|
||||
# about it being out of date.
|
||||
- "pip install --disable-pip-version-check --user --upgrade pip"
|
||||
|
||||
# Install the build dependencies of the project. If some dependencies contain
|
||||
# compiled extensions and are not provided as pre-built wheel packages,
|
||||
# pip will build them from source using the MSVC compiler matching the
|
||||
# target Python version and architecture
|
||||
- "%CMD_IN_ENV% pip install -r requirements.txt"
|
||||
|
||||
build_script:
|
||||
# Build the compiled extension
|
||||
- "%CMD_IN_ENV% python setup.py build"
|
||||
|
||||
test_script:
|
||||
# Run the project tests
|
||||
- "%CMD_IN_ENV% python setup.py test"
|
||||
|
||||
after_test:
|
||||
# If tests are successful, create binary packages for the project.
|
||||
- "%CMD_IN_ENV% python setup.py bdist_wheel"
|
||||
- "%CMD_IN_ENV% python setup.py bdist_wininst"
|
||||
- "%CMD_IN_ENV% python setup.py bdist_msi"
|
||||
- ps: "ls dist"
|
||||
|
||||
artifacts:
|
||||
# Archive the generated packages in the ci.appveyor.com build report.
|
||||
- path: dist\*
|
||||
|
||||
#on_success:
|
||||
# - TODO: upload the content of dist/*.whl to a public wheelhouse
|
||||
#
|
@ -1 +0,0 @@
|
||||
collect_ignore = ["setup.py", "build_all.py"]
|
@ -1,177 +0,0 @@
|
||||
# Makefile for Sphinx documentation
|
||||
#
|
||||
|
||||
# You can set these variables from the command line.
|
||||
SPHINXOPTS =
|
||||
SPHINXBUILD = sphinx-build
|
||||
PAPER =
|
||||
BUILDDIR = _build
|
||||
|
||||
# User-friendly check for sphinx-build
|
||||
ifeq ($(shell which $(SPHINXBUILD) >/dev/null 2>&1; echo $?), 1)
|
||||
$(error The '$(SPHINXBUILD)' command was not found. Make sure you have Sphinx installed, then set the SPHINXBUILD environment variable to point to the full path of the '$(SPHINXBUILD)' executable. Alternatively you can add the directory with the executable to your PATH. If you don't have Sphinx installed, grab it from http://sphinx-doc.org/)
|
||||
endif
|
||||
|
||||
# Internal variables.
|
||||
PAPEROPT_a4 = -D latex_paper_size=a4
|
||||
PAPEROPT_letter = -D latex_paper_size=letter
|
||||
ALLSPHINXOPTS = -d $(BUILDDIR)/doctrees $(PAPEROPT_$(PAPER)) $(SPHINXOPTS) .
|
||||
# the i18n builder cannot share the environment and doctrees with the others
|
||||
I18NSPHINXOPTS = $(PAPEROPT_$(PAPER)) $(SPHINXOPTS) .
|
||||
|
||||
.PHONY: help clean html dirhtml singlehtml pickle json htmlhelp qthelp devhelp epub latex latexpdf text man changes linkcheck doctest gettext
|
||||
|
||||
help:
|
||||
@echo "Please use \`make <target>' where <target> is one of"
|
||||
@echo " html to make standalone HTML files"
|
||||
@echo " dirhtml to make HTML files named index.html in directories"
|
||||
@echo " singlehtml to make a single large HTML file"
|
||||
@echo " pickle to make pickle files"
|
||||
@echo " json to make JSON files"
|
||||
@echo " htmlhelp to make HTML files and a HTML help project"
|
||||
@echo " qthelp to make HTML files and a qthelp project"
|
||||
@echo " devhelp to make HTML files and a Devhelp project"
|
||||
@echo " epub to make an epub"
|
||||
@echo " latex to make LaTeX files, you can set PAPER=a4 or PAPER=letter"
|
||||
@echo " latexpdf to make LaTeX files and run them through pdflatex"
|
||||
@echo " latexpdfja to make LaTeX files and run them through platex/dvipdfmx"
|
||||
@echo " text to make text files"
|
||||
@echo " man to make manual pages"
|
||||
@echo " texinfo to make Texinfo files"
|
||||
@echo " info to make Texinfo files and run them through makeinfo"
|
||||
@echo " gettext to make PO message catalogs"
|
||||
@echo " changes to make an overview of all changed/added/deprecated items"
|
||||
@echo " xml to make Docutils-native XML files"
|
||||
@echo " pseudoxml to make pseudoxml-XML files for display purposes"
|
||||
@echo " linkcheck to check all external links for integrity"
|
||||
@echo " doctest to run all doctests embedded in the documentation (if enabled)"
|
||||
|
||||
clean:
|
||||
rm -rf $(BUILDDIR)/*
|
||||
|
||||
html:
|
||||
$(SPHINXBUILD) -b html $(ALLSPHINXOPTS) $(BUILDDIR)/html
|
||||
@echo
|
||||
@echo "Build finished. The HTML pages are in $(BUILDDIR)/html."
|
||||
|
||||
dirhtml:
|
||||
$(SPHINXBUILD) -b dirhtml $(ALLSPHINXOPTS) $(BUILDDIR)/dirhtml
|
||||
@echo
|
||||
@echo "Build finished. The HTML pages are in $(BUILDDIR)/dirhtml."
|
||||
|
||||
singlehtml:
|
||||
$(SPHINXBUILD) -b singlehtml $(ALLSPHINXOPTS) $(BUILDDIR)/singlehtml
|
||||
@echo
|
||||
@echo "Build finished. The HTML page is in $(BUILDDIR)/singlehtml."
|
||||
|
||||
pickle:
|
||||
$(SPHINXBUILD) -b pickle $(ALLSPHINXOPTS) $(BUILDDIR)/pickle
|
||||
@echo
|
||||
@echo "Build finished; now you can process the pickle files."
|
||||
|
||||
json:
|
||||
$(SPHINXBUILD) -b json $(ALLSPHINXOPTS) $(BUILDDIR)/json
|
||||
@echo
|
||||
@echo "Build finished; now you can process the JSON files."
|
||||
|
||||
htmlhelp:
|
||||
$(SPHINXBUILD) -b htmlhelp $(ALLSPHINXOPTS) $(BUILDDIR)/htmlhelp
|
||||
@echo
|
||||
@echo "Build finished; now you can run HTML Help Workshop with the" \
|
||||
".hhp project file in $(BUILDDIR)/htmlhelp."
|
||||
|
||||
qthelp:
|
||||
$(SPHINXBUILD) -b qthelp $(ALLSPHINXOPTS) $(BUILDDIR)/qthelp
|
||||
@echo
|
||||
@echo "Build finished; now you can run "qcollectiongenerator" with the" \
|
||||
".qhcp project file in $(BUILDDIR)/qthelp, like this:"
|
||||
@echo "# qcollectiongenerator $(BUILDDIR)/qthelp/pywafo.qhcp"
|
||||
@echo "To view the help file:"
|
||||
@echo "# assistant -collectionFile $(BUILDDIR)/qthelp/pywafo.qhc"
|
||||
|
||||
devhelp:
|
||||
$(SPHINXBUILD) -b devhelp $(ALLSPHINXOPTS) $(BUILDDIR)/devhelp
|
||||
@echo
|
||||
@echo "Build finished."
|
||||
@echo "To view the help file:"
|
||||
@echo "# mkdir -p $HOME/.local/share/devhelp/pywafo"
|
||||
@echo "# ln -s $(BUILDDIR)/devhelp $HOME/.local/share/devhelp/pywafo"
|
||||
@echo "# devhelp"
|
||||
|
||||
epub:
|
||||
$(SPHINXBUILD) -b epub $(ALLSPHINXOPTS) $(BUILDDIR)/epub
|
||||
@echo
|
||||
@echo "Build finished. The epub file is in $(BUILDDIR)/epub."
|
||||
|
||||
latex:
|
||||
$(SPHINXBUILD) -b latex $(ALLSPHINXOPTS) $(BUILDDIR)/latex
|
||||
@echo
|
||||
@echo "Build finished; the LaTeX files are in $(BUILDDIR)/latex."
|
||||
@echo "Run \`make' in that directory to run these through (pdf)latex" \
|
||||
"(use \`make latexpdf' here to do that automatically)."
|
||||
|
||||
latexpdf:
|
||||
$(SPHINXBUILD) -b latex $(ALLSPHINXOPTS) $(BUILDDIR)/latex
|
||||
@echo "Running LaTeX files through pdflatex..."
|
||||
$(MAKE) -C $(BUILDDIR)/latex all-pdf
|
||||
@echo "pdflatex finished; the PDF files are in $(BUILDDIR)/latex."
|
||||
|
||||
latexpdfja:
|
||||
$(SPHINXBUILD) -b latex $(ALLSPHINXOPTS) $(BUILDDIR)/latex
|
||||
@echo "Running LaTeX files through platex and dvipdfmx..."
|
||||
$(MAKE) -C $(BUILDDIR)/latex all-pdf-ja
|
||||
@echo "pdflatex finished; the PDF files are in $(BUILDDIR)/latex."
|
||||
|
||||
text:
|
||||
$(SPHINXBUILD) -b text $(ALLSPHINXOPTS) $(BUILDDIR)/text
|
||||
@echo
|
||||
@echo "Build finished. The text files are in $(BUILDDIR)/text."
|
||||
|
||||
man:
|
||||
$(SPHINXBUILD) -b man $(ALLSPHINXOPTS) $(BUILDDIR)/man
|
||||
@echo
|
||||
@echo "Build finished. The manual pages are in $(BUILDDIR)/man."
|
||||
|
||||
texinfo:
|
||||
$(SPHINXBUILD) -b texinfo $(ALLSPHINXOPTS) $(BUILDDIR)/texinfo
|
||||
@echo
|
||||
@echo "Build finished. The Texinfo files are in $(BUILDDIR)/texinfo."
|
||||
@echo "Run \`make' in that directory to run these through makeinfo" \
|
||||
"(use \`make info' here to do that automatically)."
|
||||
|
||||
info:
|
||||
$(SPHINXBUILD) -b texinfo $(ALLSPHINXOPTS) $(BUILDDIR)/texinfo
|
||||
@echo "Running Texinfo files through makeinfo..."
|
||||
make -C $(BUILDDIR)/texinfo info
|
||||
@echo "makeinfo finished; the Info files are in $(BUILDDIR)/texinfo."
|
||||
|
||||
gettext:
|
||||
$(SPHINXBUILD) -b gettext $(I18NSPHINXOPTS) $(BUILDDIR)/locale
|
||||
@echo
|
||||
@echo "Build finished. The message catalogs are in $(BUILDDIR)/locale."
|
||||
|
||||
changes:
|
||||
$(SPHINXBUILD) -b changes $(ALLSPHINXOPTS) $(BUILDDIR)/changes
|
||||
@echo
|
||||
@echo "The overview file is in $(BUILDDIR)/changes."
|
||||
|
||||
linkcheck:
|
||||
$(SPHINXBUILD) -b linkcheck $(ALLSPHINXOPTS) $(BUILDDIR)/linkcheck
|
||||
@echo
|
||||
@echo "Link check complete; look for any errors in the above output " \
|
||||
"or in $(BUILDDIR)/linkcheck/output.txt."
|
||||
|
||||
doctest:
|
||||
$(SPHINXBUILD) -b doctest $(ALLSPHINXOPTS) $(BUILDDIR)/doctest
|
||||
@echo "Testing of doctests in the sources finished, look at the " \
|
||||
"results in $(BUILDDIR)/doctest/output.txt."
|
||||
|
||||
xml:
|
||||
$(SPHINXBUILD) -b xml $(ALLSPHINXOPTS) $(BUILDDIR)/xml
|
||||
@echo
|
||||
@echo "Build finished. The XML files are in $(BUILDDIR)/xml."
|
||||
|
||||
pseudoxml:
|
||||
$(SPHINXBUILD) -b pseudoxml $(ALLSPHINXOPTS) $(BUILDDIR)/pseudoxml
|
||||
@echo
|
||||
@echo "Build finished. The pseudo-XML files are in $(BUILDDIR)/pseudoxml."
|
@ -1,2 +0,0 @@
|
||||
.. _authors:
|
||||
.. include:: ../AUTHORS.rst
|
@ -1,2 +0,0 @@
|
||||
.. _changes:
|
||||
.. include:: ../CHANGES.rst
|
@ -1,238 +0,0 @@
|
||||
# -*- coding: utf-8 -*-
|
||||
#
|
||||
# This file is execfile()d with the current directory set to its containing dir.
|
||||
#
|
||||
# Note that not all possible configuration values are present in this
|
||||
# autogenerated file.
|
||||
#
|
||||
# All configuration values have a default; values that are commented out
|
||||
# serve to show the default.
|
||||
|
||||
import sys
|
||||
import os
|
||||
import inspect
|
||||
from sphinx import apidoc
|
||||
|
||||
|
||||
__location__ = os.path.join(os.getcwd(), os.path.dirname(
|
||||
inspect.getfile(inspect.currentframe())))
|
||||
|
||||
package = "wafo"
|
||||
namespace = []
|
||||
namespace_pkg = ".".join([namespace[-1], package]) if namespace else package
|
||||
|
||||
# If extensions (or modules to document with autodoc) are in another directory,
|
||||
# add these directories to sys.path here. If the directory is relative to the
|
||||
# documentation root, use os.path.abspath to make it absolute, like shown here.
|
||||
# sys.path.insert(0, os.path.abspath('.'))
|
||||
|
||||
# -- General configuration -----------------------------------------------------
|
||||
|
||||
# If your documentation needs a minimal Sphinx version, state it here.
|
||||
# needs_sphinx = '1.0'
|
||||
|
||||
# Add any Sphinx extension module names here, as strings. They can be extensions
|
||||
# coming with Sphinx (named 'sphinx.ext.*') or your custom ones.
|
||||
extensions = ['sphinx.ext.autodoc', 'sphinx.ext.intersphinx', 'sphinx.ext.todo',
|
||||
'sphinx.ext.autosummary', 'sphinx.ext.viewcode', 'sphinx.ext.coverage',
|
||||
'sphinx.ext.doctest', 'sphinx.ext.ifconfig', 'sphinx.ext.pngmath']
|
||||
|
||||
# Add any paths that contain templates here, relative to this directory.
|
||||
templates_path = ['_templates']
|
||||
|
||||
# The suffix of source filenames.
|
||||
source_suffix = '.rst'
|
||||
|
||||
# The encoding of source files.
|
||||
# source_encoding = 'utf-8-sig'
|
||||
|
||||
# The master toctree document.
|
||||
master_doc = 'index'
|
||||
|
||||
# General information about the project.
|
||||
project = u'pywafo'
|
||||
copyright = u'2015, davidovitch'
|
||||
|
||||
# The version info for the project you're documenting, acts as replacement for
|
||||
# |version| and |release|, also used in various other places throughout the
|
||||
# built documents.
|
||||
#
|
||||
# The short X.Y version.
|
||||
version = '' # Is set by calling `setup.py docs`
|
||||
# The full version, including alpha/beta/rc tags.
|
||||
release = '' # Is set by calling `setup.py docs`
|
||||
|
||||
# The language for content autogenerated by Sphinx. Refer to documentation
|
||||
# for a list of supported languages.
|
||||
# language = None
|
||||
|
||||
# There are two options for replacing |today|: either, you set today to some
|
||||
# non-false value, then it is used:
|
||||
# today = ''
|
||||
# Else, today_fmt is used as the format for a strftime call.
|
||||
# today_fmt = '%B %d, %Y'
|
||||
|
||||
# List of patterns, relative to source directory, that match files and
|
||||
# directories to ignore when looking for source files.
|
||||
exclude_patterns = ['_build']
|
||||
|
||||
# The reST default role (used for this markup: `text`) to use for all documents.
|
||||
# default_role = None
|
||||
|
||||
# If true, '()' will be appended to :func: etc. cross-reference text.
|
||||
# add_function_parentheses = True
|
||||
|
||||
# If true, the current module name will be prepended to all description
|
||||
# unit titles (such as .. function::).
|
||||
# add_module_names = True
|
||||
|
||||
# If true, sectionauthor and moduleauthor directives will be shown in the
|
||||
# output. They are ignored by default.
|
||||
# show_authors = False
|
||||
|
||||
# The name of the Pygments (syntax highlighting) style to use.
|
||||
pygments_style = 'sphinx'
|
||||
|
||||
# A list of ignored prefixes for module index sorting.
|
||||
# modindex_common_prefix = []
|
||||
|
||||
# If true, keep warnings as "system message" paragraphs in the built documents.
|
||||
# keep_warnings = False
|
||||
|
||||
|
||||
# -- Options for HTML output ---------------------------------------------------
|
||||
|
||||
# The theme to use for HTML and HTML Help pages. See the documentation for
|
||||
# a list of builtin themes.
|
||||
html_theme = 'alabaster'
|
||||
|
||||
# Theme options are theme-specific and customize the look and feel of a theme
|
||||
# further. For a list of options available for each theme, see the
|
||||
# documentation.
|
||||
# html_theme_options = {}
|
||||
|
||||
# Add any paths that contain custom themes here, relative to this directory.
|
||||
# html_theme_path = []
|
||||
|
||||
# The name for this set of Sphinx documents. If None, it defaults to
|
||||
# "<project> v<release> documentation".
|
||||
try:
|
||||
from namespace_pkg import __version__ as version
|
||||
except ImportError:
|
||||
pass
|
||||
else:
|
||||
release = version
|
||||
|
||||
# A shorter title for the navigation bar. Default is the same as html_title.
|
||||
# html_short_title = None
|
||||
|
||||
# The name of an image file (relative to this directory) to place at the top
|
||||
# of the sidebar.
|
||||
# html_logo = ""
|
||||
|
||||
# The name of an image file (within the static path) to use as favicon of the
|
||||
# docs. This file should be a Windows icon file (.ico) being 16x16 or 32x32
|
||||
# pixels large.
|
||||
# html_favicon = None
|
||||
|
||||
# Add any paths that contain custom static files (such as style sheets) here,
|
||||
# relative to this directory. They are copied after the builtin static files,
|
||||
# so a file named "default.css" will overwrite the builtin "default.css".
|
||||
html_static_path = ['_static']
|
||||
|
||||
# If not '', a 'Last updated on:' timestamp is inserted at every page bottom,
|
||||
# using the given strftime format.
|
||||
# html_last_updated_fmt = '%b %d, %Y'
|
||||
|
||||
# If true, SmartyPants will be used to convert quotes and dashes to
|
||||
# typographically correct entities.
|
||||
# html_use_smartypants = True
|
||||
|
||||
# Custom sidebar templates, maps document names to template names.
|
||||
# html_sidebars = {}
|
||||
|
||||
# Additional templates that should be rendered to pages, maps page names to
|
||||
# template names.
|
||||
# html_additional_pages = {}
|
||||
|
||||
# If false, no module index is generated.
|
||||
# html_domain_indices = True
|
||||
|
||||
# If false, no index is generated.
|
||||
# html_use_index = True
|
||||
|
||||
# If true, the index is split into individual pages for each letter.
|
||||
# html_split_index = False
|
||||
|
||||
# If true, links to the reST sources are added to the pages.
|
||||
# html_show_sourcelink = True
|
||||
|
||||
# If true, "Created using Sphinx" is shown in the HTML footer. Default is True.
|
||||
# html_show_sphinx = True
|
||||
|
||||
# If true, "(C) Copyright ..." is shown in the HTML footer. Default is True.
|
||||
# html_show_copyright = True
|
||||
|
||||
# If true, an OpenSearch description file will be output, and all pages will
|
||||
# contain a <link> tag referring to it. The value of this option must be the
|
||||
# base URL from which the finished HTML is served.
|
||||
# html_use_opensearch = ''
|
||||
|
||||
# This is the file name suffix for HTML files (e.g. ".xhtml").
|
||||
# html_file_suffix = None
|
||||
|
||||
# Output file base name for HTML help builder.
|
||||
htmlhelp_basename = 'wafo-doc'
|
||||
|
||||
|
||||
# -- Options for LaTeX output --------------------------------------------------
|
||||
|
||||
latex_elements = {
|
||||
# The paper size ('letterpaper' or 'a4paper').
|
||||
# 'papersize': 'letterpaper',
|
||||
|
||||
# The font size ('10pt', '11pt' or '12pt').
|
||||
# 'pointsize': '10pt',
|
||||
|
||||
# Additional stuff for the LaTeX preamble.
|
||||
# 'preamble': '',
|
||||
}
|
||||
|
||||
# Grouping the document tree into LaTeX files. List of tuples
|
||||
# (source start file, target name, title, author, documentclass [howto/manual]).
|
||||
latex_documents = [
|
||||
('index', 'user_guide.tex', u'pywafo Documentation',
|
||||
u'davidovitch', 'manual'),
|
||||
]
|
||||
|
||||
# The name of an image file (relative to this directory) to place at the top of
|
||||
# the title page.
|
||||
# latex_logo = ""
|
||||
|
||||
# For "manual" documents, if this is true, then toplevel headings are parts,
|
||||
# not chapters.
|
||||
# latex_use_parts = False
|
||||
|
||||
# If true, show page references after internal links.
|
||||
# latex_show_pagerefs = False
|
||||
|
||||
# If true, show URL addresses after external links.
|
||||
# latex_show_urls = False
|
||||
|
||||
# Documents to append as an appendix to all manuals.
|
||||
# latex_appendices = []
|
||||
|
||||
# If false, no module index is generated.
|
||||
# latex_domain_indices = True
|
||||
|
||||
# -- External mapping ------------------------------------------------------------
|
||||
python_version = '.'.join(map(str, sys.version_info[0:2]))
|
||||
intersphinx_mapping = {
|
||||
'sphinx': ('http://sphinx.pocoo.org', None),
|
||||
'python': ('http://docs.python.org/' + python_version, None),
|
||||
'matplotlib': ('http://matplotlib.sourceforge.net', None),
|
||||
'numpy': ('http://docs.scipy.org/doc/numpy', None),
|
||||
'sklearn': ('http://scikit-learn.org/stable', None),
|
||||
'pandas': ('http://pandas.pydata.org/pandas-docs/stable', None),
|
||||
'scipy': ('http://docs.scipy.org/doc/scipy/reference/', None),
|
||||
}
|
@ -1,45 +0,0 @@
|
||||
======
|
||||
pywafo
|
||||
======
|
||||
|
||||
This is the documentation of **pywafo**.
|
||||
|
||||
.. note::
|
||||
|
||||
This is the main page of your project's `Sphinx <http://sphinx-doc.org/>`_
|
||||
documentation. It is formatted in `reStructuredText
|
||||
<http://sphinx-doc.org/rest.html>`__. Add additional pages by creating
|
||||
rst-files in ``docs`` and adding them to the `toctree
|
||||
<http://sphinx-doc.org/markup/toctree.html>`_ below. Use then
|
||||
`references <http://sphinx-doc.org/markup/inline.html>`__ in order to link
|
||||
them from this page, e.g. :ref:`authors <authors>` and :ref:`changes`.
|
||||
It is also possible to refer to the documentation of other Python packages
|
||||
with the `Python domain syntax
|
||||
<http://sphinx-doc.org/domains.html#the-python-domain>`__. By default you
|
||||
can reference the documentation of `Sphinx <http://sphinx.pocoo.org>`__,
|
||||
`Python <http://docs.python.org/>`__, `matplotlib
|
||||
<http://matplotlib.sourceforge.net>`__, `NumPy
|
||||
<http://docs.scipy.org/doc/numpy>`__, `Scikit-Learn
|
||||
<http://scikit-learn.org/stable>`__, `Pandas
|
||||
<http://pandas.pydata.org/pandas-docs/stable>`__, `SciPy
|
||||
<http://docs.scipy.org/doc/scipy/reference/>`__. You can add more by
|
||||
extending the ``intersphinx_mapping`` in your Sphinx's ``conf.py``.
|
||||
|
||||
Contents
|
||||
========
|
||||
|
||||
.. toctree::
|
||||
:maxdepth: 2
|
||||
|
||||
License <license>
|
||||
Authors <authors>
|
||||
Changelog <changes>
|
||||
Module Reference <api/modules>
|
||||
|
||||
|
||||
Indices and tables
|
||||
==================
|
||||
|
||||
* :ref:`genindex`
|
||||
* :ref:`modindex`
|
||||
* :ref:`search`
|
@ -1,7 +0,0 @@
|
||||
.. _license:
|
||||
|
||||
=======
|
||||
License
|
||||
=======
|
||||
|
||||
.. literalinclude:: ../LICENSE.txt
|
@ -1,110 +1,10 @@
|
||||
#!/usr/bin/env python
|
||||
# -*- coding: utf-8 -*-
|
||||
"""
|
||||
Setup file for wafo.
|
||||
|
||||
This file was generated with PyScaffold 2.4.2, a tool that easily
|
||||
puts up a scaffold for your new Python project. Learn more under:
|
||||
http://pyscaffold.readthedocs.org/
|
||||
|
||||
Usage:
|
||||
Run all tests:
|
||||
python setup.py test
|
||||
|
||||
python setup.py doctests
|
||||
|
||||
Build documentation
|
||||
|
||||
python setup.py docs
|
||||
|
||||
Install
|
||||
python setup.py install [, --prefix=$PREFIX]
|
||||
|
||||
Build
|
||||
|
||||
python setup.py bdist_wininst
|
||||
|
||||
python setup.py bdist_wheel --universal
|
||||
|
||||
python setup.py sdist
|
||||
|
||||
PyPi upload:
|
||||
twine upload dist/*
|
||||
|
||||
"""
|
||||
|
||||
from __future__ import division, absolute_import, print_function
|
||||
|
||||
# numpy.distutils will figure out if setuptools is available when imported
|
||||
# this allows us to combine setuptools use_pyscaffold=True and f2py extensions
|
||||
import setuptools
|
||||
from numpy.distutils.core import setup
|
||||
from numpy.distutils.misc_util import Configuration
|
||||
|
||||
import sys
|
||||
|
||||
|
||||
def setup_package_pyscaffold():
|
||||
|
||||
config = Configuration('wafo')
|
||||
|
||||
# -------------------------------------------------------------------------
|
||||
# c_library
|
||||
config.add_extension('c_library',
|
||||
sources=['wafo/source/c_library/c_library.pyf',
|
||||
'wafo/source/c_library/c_functions.c'])
|
||||
# -------------------------------------------------------------------------
|
||||
# mvn
|
||||
config.add_extension('mvn',
|
||||
sources=['wafo/source/mvn/mvn.pyf',
|
||||
'wafo/source/mvn/mvndst.f'])
|
||||
|
||||
# -------------------------------------------------------------------------
|
||||
# mvnprdmod
|
||||
lib_mvnprdmod_src = ['wafo/source/mvnprd/mvnprd.f',
|
||||
'wafo/source/mvnprd/mvnprodcorrprb.f']
|
||||
config.add_library('_mvnprdmod', sources=lib_mvnprdmod_src)
|
||||
config.add_extension('mvnprdmod',
|
||||
sources=['wafo/source/mvnprd/mvnprd_interface.f'],
|
||||
libraries=['_mvnprdmod'],
|
||||
depends=(lib_mvnprdmod_src))
|
||||
|
||||
# -------------------------------------------------------------------------
|
||||
# cov2mod
|
||||
lib_cov2mod_src = ['wafo/source/mreg/dsvdc.f',
|
||||
'wafo/source/mreg/mregmodule.f',
|
||||
'wafo/source/mreg/intfcmod.f']
|
||||
config.add_library('_cov2mod', sources=lib_cov2mod_src)
|
||||
config.add_extension('cov2mod',
|
||||
sources=['wafo/source/mreg/cov2mmpdfreg_intfc.f'],
|
||||
libraries=['_cov2mod'],
|
||||
include_dirs=['wafo/source/mreg/'],
|
||||
depends=(lib_cov2mod_src))
|
||||
|
||||
# -------------------------------------------------------------------------
|
||||
# rindmod
|
||||
lib_rindmod_src = ['wafo/source/rind2007/intmodule.f',
|
||||
'wafo/source/rind2007/jacobmod.f',
|
||||
'wafo/source/rind2007/swapmod.f',
|
||||
'wafo/source/rind2007/fimod.f',
|
||||
'wafo/source/rind2007/rindmod.f',
|
||||
'wafo/source/rind2007/rind71mod.f']
|
||||
config.add_library('_rindmod', sources=lib_rindmod_src)
|
||||
config.add_extension('rindmod',
|
||||
sources=['wafo/source/rind2007/rind_interface.f'],
|
||||
libraries=['_rindmod'],
|
||||
include_dirs=['wafo/source/mreg/'],
|
||||
depends=(lib_rindmod_src))
|
||||
|
||||
config.add_data_dir(('data', 'wafo/data'))
|
||||
|
||||
needs_sphinx = {'build_sphinx', 'upload_docs'}.intersection(sys.argv)
|
||||
sphinx = ['sphinx'] if needs_sphinx else []
|
||||
setup(setup_requires=['six', 'pyscaffold>=2.4rc1,<2.5a0'] + sphinx,
|
||||
tests_require=['pytest_cov', 'pytest'],
|
||||
use_pyscaffold=True,
|
||||
**config.todict())
|
||||
|
||||
|
||||
if __name__ == "__main__":
|
||||
setup_package_pyscaffold()
|
||||
setuptools.setup(
|
||||
name='wafo',
|
||||
version='0.11',
|
||||
author='WAFO-group',
|
||||
author_email=None,
|
||||
description='Wave Analysis for Fatigue and Oceanography',
|
||||
packages=setuptools.find_packages(),
|
||||
)
|
||||
|
@ -1,69 +0,0 @@
|
||||
<map version="0.8.1">
|
||||
<!-- To view this file, download free mind mapping software FreeMind from http://freemind.sourceforge.net -->
|
||||
<node CREATED="1268907205249" ID="Freemind_Link_1834025809" MODIFIED="1268920224244" STYLE="fork" TEXT="WAFO Objects">
|
||||
<font NAME="SansSerif" SIZE="11"/>
|
||||
<node CREATED="1268919537155" ID="Freemind_Link_823244205" MODIFIED="1268920190865" POSITION="right" STYLE="bubble" TEXT="SpecData1D" VSHIFT="36">
|
||||
<node CREATED="1268908695254" ID="Freemind_Link_834391422" MODIFIED="1268908707582" TEXT="bandwidth"/>
|
||||
<node CREATED="1268908733082" ID="Freemind_Link_1433599166" MODIFIED="1268908736348" TEXT="characteristic"/>
|
||||
<node CREATED="1268908690441" ID="Freemind_Link_1284342306" MODIFIED="1268908694941" TEXT="normalize"/>
|
||||
<node CREATED="1268908669144" ID="Freemind_Link_1815459961" MODIFIED="1268908677316" TEXT="nyquist_freq"/>
|
||||
<node CREATED="1268908664160" ID="Freemind_Link_703739784" MODIFIED="1268919800253" TEXT="moment" VSHIFT="4"/>
|
||||
<node CREATED="1268908678394" ID="Freemind_Link_1387269779" MODIFIED="1268919203464" TEXT="sampling_period" VSHIFT="3"/>
|
||||
<node CREATED="1268908643159" ID="Freemind_Link_895051601" MODIFIED="1268908655410" TEXT="stats_nl"/>
|
||||
<node CREATED="1268907362986" ID="Freemind_Link_1959199999" MODIFIED="1268920763548" TEXT="sim">
|
||||
<arrowlink DESTINATION="Freemind_Link_1110096845" ENDARROW="Default" ENDINCLINATION="-26;-470;" ID="Freemind_Arrow_Link_658989110" STARTARROW="None" STARTINCLINATION="134;20;"/>
|
||||
</node>
|
||||
<node CREATED="1268908639894" ID="Freemind_Link_881848843" MODIFIED="1268920768236" TEXT="sim_nl" VSHIFT="-116">
|
||||
<arrowlink DESTINATION="Freemind_Link_1110096845" ENDARROW="Default" ENDINCLINATION="-12;-203;" ID="Freemind_Arrow_Link_194892354" STARTARROW="None" STARTINCLINATION="286;65;"/>
|
||||
</node>
|
||||
<node CREATED="1268908656097" ID="Freemind_Link_932055041" MODIFIED="1268920049331" TEXT="testgaussian" VSHIFT="-11"/>
|
||||
<node CREATED="1268908524408" ID="Freemind_Link_451836141" MODIFIED="1268920637967" TEXT="tocovdata" VSHIFT="122">
|
||||
<arrowlink DESTINATION="Freemind_Link_1298097150" ENDARROW="Default" ENDINCLINATION="22;-61;" ID="Freemind_Arrow_Link_459039584" STARTARROW="None" STARTINCLINATION="155;-17;"/>
|
||||
</node>
|
||||
<node CREATED="1268907881846" ID="Freemind_Link_961675105" MODIFIED="1268908523861" TEXT="tocov_matrix"/>
|
||||
<node CREATED="1268908531877" ID="Freemind_Link_1361566099" MODIFIED="1268920224244" TEXT="to_linspec"/>
|
||||
<node CREATED="1268908548674" ID="Freemind_Link_28901814" MODIFIED="1268920222025" TEXT="to_specnorm"/>
|
||||
<node CREATED="1268908684613" ID="Freemind_Link_180569473" MODIFIED="1268917666022" TEXT="resample"/>
|
||||
</node>
|
||||
<node CREATED="1268918310233" HGAP="471" ID="Freemind_Link_1110096845" MODIFIED="1268920768236" POSITION="right" STYLE="bubble" TEXT="TimeSeries" VSHIFT="-65">
|
||||
<node CREATED="1268918758255" ID="Freemind_Link_1306080198" MODIFIED="1268920381237" STYLE="fork" TEXT="tospecdata"/>
|
||||
<node CREATED="1268918767833" ID="Freemind_Link_1037382917" MODIFIED="1268920384363" STYLE="fork" TEXT="tocovdata"/>
|
||||
<node CREATED="1268918809021" ID="Freemind_Link_767625747" MODIFIED="1268918816990" STYLE="fork" TEXT="wave_periods"/>
|
||||
<node CREATED="1268918817396" ID="Freemind_Link_294440483" MODIFIED="1268918823943" STYLE="fork" TEXT="reconstruct"/>
|
||||
<node CREATED="1268918824365" ID="Freemind_Link_1893872699" MODIFIED="1268920714281" STYLE="fork" TEXT="plot_wave" VSHIFT="29"/>
|
||||
<node CREATED="1268918835318" ID="Freemind_Link_1625232632" MODIFIED="1268918848897" STYLE="fork" TEXT="plot_sp_wave"/>
|
||||
<node CREATED="1268918669644" ID="Freemind_Link_525365373" MODIFIED="1268918757817" STYLE="fork" TEXT="sampling_period"/>
|
||||
<node CREATED="1268918800427" ID="Freemind_Link_540262488" MODIFIED="1268920657436" STYLE="fork" TEXT="through_crest" VSHIFT="11">
|
||||
<arrowlink DESTINATION="Freemind_Link_1866699423" ENDARROW="Default" ENDINCLINATION="120;-87;" ID="Freemind_Arrow_Link_724369012" STARTARROW="None" STARTINCLINATION="298;0;"/>
|
||||
</node>
|
||||
<node CREATED="1268918781115" ID="Freemind_Link_191369931" MODIFIED="1268920652670" STYLE="fork" TEXT="turning_points">
|
||||
<arrowlink DESTINATION="Freemind_Link_1866699423" ENDARROW="Default" ENDINCLINATION="354;-263;" ID="Freemind_Arrow_Link_301423490" STARTARROW="None" STARTINCLINATION="-57;21;"/>
|
||||
</node>
|
||||
</node>
|
||||
<node CREATED="1268917687085" ID="Freemind_Link_1298097150" MODIFIED="1268920673171" POSITION="right" STYLE="bubble" TEXT="CovData1D" VSHIFT="-212">
|
||||
<node CREATED="1268917901666" ID="Freemind_Link_1791940512" MODIFIED="1268919887910" STYLE="fork" TEXT="tospecdata">
|
||||
<arrowlink DESTINATION="Freemind_Link_823244205" ENDARROW="Default" ENDINCLINATION="14;87;" ID="Freemind_Arrow_Link_728651707" STARTARROW="None" STARTINCLINATION="-21;-38;"/>
|
||||
</node>
|
||||
<node CREATED="1268918164841" ID="Freemind_Link_1718380955" MODIFIED="1268918173622" STYLE="fork" TEXT="sampling_period"/>
|
||||
<node CREATED="1268918174044" ID="Freemind_Link_1888500587" MODIFIED="1268920740704" STYLE="fork" TEXT="sim">
|
||||
<arrowlink DESTINATION="Freemind_Link_1110096845" ENDARROW="Default" ENDINCLINATION="-2;21;" ID="Freemind_Arrow_Link_174025984" STARTARROW="None" STARTINCLINATION="271;12;"/>
|
||||
</node>
|
||||
<node CREATED="1268918178013" ID="Freemind_Link_757742844" MODIFIED="1268920750954" STYLE="fork" TEXT="simcond">
|
||||
<arrowlink DESTINATION="Freemind_Link_1110096845" ENDARROW="Default" ENDINCLINATION="-46;173;" ID="Freemind_Arrow_Link_1889053469" STARTARROW="None" STARTINCLINATION="171;0;"/>
|
||||
</node>
|
||||
</node>
|
||||
<node CREATED="1268918904132" ID="Freemind_Link_1866699423" MODIFIED="1268920673171" POSITION="right" STYLE="bubble" TEXT="TurningPoints">
|
||||
<node CREATED="1268918978226" ID="Freemind_Link_1033172969" MODIFIED="1268918989023" STYLE="fork" TEXT="cycle_pairs"/>
|
||||
</node>
|
||||
<node CREATED="1268918988977" ID="Freemind_Link_280025207" MODIFIED="1268920787861" POSITION="right" STYLE="bubble" TEXT="CyclePairs" VSHIFT="91">
|
||||
<node CREATED="1268919011477" ID="Freemind_Link_867889153" MODIFIED="1268919019180" STYLE="fork" TEXT="amplitudes"/>
|
||||
<node CREATED="1268919019586" ID="Freemind_Link_40128767" MODIFIED="1268919024414" STYLE="fork" TEXT="damage"/>
|
||||
<node CREATED="1268919025024" ID="Freemind_Link_1194799087" MODIFIED="1268919033727" STYLE="fork" TEXT="level_crossings"/>
|
||||
</node>
|
||||
<node CREATED="1268919295668" ID="Freemind_Link_106637151" MODIFIED="1268919305606" POSITION="right" STYLE="bubble" TEXT="LevelCrossings">
|
||||
<node CREATED="1268919349809" ID="Freemind_Link_1941466065" MODIFIED="1268920453960" STYLE="fork" TEXT="sim"/>
|
||||
<node CREATED="1268919305574" ID="Freemind_Link_1378561112" MODIFIED="1268919349122" STYLE="fork" TEXT="trdata"/>
|
||||
</node>
|
||||
<node CREATED="1268920796033" ID="_" MODIFIED="1268920800424" POSITION="right" STYLE="bubble" TEXT="TrData"/>
|
||||
</node>
|
||||
</map>
|
Binary file not shown.
Before Width: | Height: | Size: 240 KiB |
@ -1,97 +0,0 @@
|
||||
"""
|
||||
Module extending the bitoperator capabilites of numpy
|
||||
"""
|
||||
|
||||
from numpy import (bitwise_and, bitwise_or,
|
||||
bitwise_not, binary_repr, # @UnusedImport
|
||||
bitwise_xor, where, arange) # @UnusedImport
|
||||
__all__ = ['bitwise_and', 'bitwise_or', 'bitwise_not', 'binary_repr',
|
||||
'bitwise_xor', 'getbit', 'setbit', 'getbits', 'setbits']
|
||||
|
||||
|
||||
def getbit(i, bit):
|
||||
"""
|
||||
Get bit at specified position
|
||||
|
||||
Parameters
|
||||
----------
|
||||
i : array-like of uints, longs
|
||||
value to
|
||||
bit : array-like of ints or longs
|
||||
bit position between 0 and the number of bits in the uint class.
|
||||
|
||||
Examples
|
||||
--------
|
||||
>>> import numpy as np
|
||||
>>> binary_repr(13)
|
||||
'1101'
|
||||
>>> getbit(13,np.arange(3,-1,-1))
|
||||
array([1, 1, 0, 1])
|
||||
>>> getbit(5, np.r_[0:4])
|
||||
array([1, 0, 1, 0])
|
||||
"""
|
||||
return bitwise_and(i, 1 << bit) >> bit
|
||||
|
||||
|
||||
def getbits(i, numbits=8):
|
||||
"""
|
||||
Returns bits of i in a list
|
||||
"""
|
||||
return getbit(i, arange(0, numbits))
|
||||
|
||||
|
||||
def setbit(i, bit, value=1):
|
||||
"""
|
||||
Set bit at specified position
|
||||
|
||||
Parameters
|
||||
----------
|
||||
i : array-like of uints, longs
|
||||
value to
|
||||
bit : array-like of ints or longs
|
||||
bit position between 0 and the number of bits in the uint class.
|
||||
value : array-like of 0 or 1
|
||||
value to set the bit to.
|
||||
|
||||
Examples
|
||||
--------
|
||||
Set bit fifth bit in the five bit binary binary representation of 9 (01001)
|
||||
yields 25 (11001)
|
||||
>>> setbit(9,4)
|
||||
array(25)
|
||||
"""
|
||||
val1 = 1 << bit
|
||||
val0 = bitwise_not(val1)
|
||||
return where((value == 0) & (i == i) & (bit == bit), bitwise_and(i, val0),
|
||||
bitwise_or(i, val1))
|
||||
|
||||
|
||||
def setbits(bitlist):
|
||||
"""
|
||||
Set bits of val to values in bitlist
|
||||
|
||||
Example
|
||||
-------
|
||||
>>> setbits([1,1])
|
||||
3
|
||||
>>> setbits([1,0])
|
||||
1
|
||||
"""
|
||||
val = 0
|
||||
for i, j in enumerate(bitlist):
|
||||
val |= j << i
|
||||
return val
|
||||
|
||||
if __name__ == '__main__':
|
||||
from wafo.testing import test_docstrings
|
||||
test_docstrings(__file__)
|
||||
|
||||
# t = set(np.arange(8),1,1)
|
||||
# t=get(0x84,np.arange(0,8))
|
||||
# t=getbyte(0x84)
|
||||
# t=get(0x84,[0, 1, 2, 3, 4, 5, 6, 7])
|
||||
# t=get(0x20, 6)
|
||||
# bit = [0 for i in range(8)]
|
||||
# bit[7]=1
|
||||
# t = setbits(bit)
|
||||
# print(hex(t))
|
@ -1,3 +0,0 @@
|
||||
# TODO Fix doctests in fig.py
|
||||
collect_ignore = ["fig.py", "MSO.py", "MSPPT.py", "powerpoint.py",
|
||||
"win32_utils.py"]
|
@ -1,3 +0,0 @@
|
||||
from __future__ import absolute_import
|
||||
from .info import __doc__
|
||||
from .info import *
|
@ -1,582 +0,0 @@
|
||||
5.4829629629629641e+00
|
||||
4.3614999999999986e+00
|
||||
5.2602325581395339e+00
|
||||
3.0619230769230770e+00
|
||||
4.4896296296296301e+00
|
||||
3.3681395348837206e+00
|
||||
4.0259523809523818e+00
|
||||
4.3966666666666665e+00
|
||||
4.2999999999999998e+00
|
||||
7.0564705882352934e+00
|
||||
6.1440000000000001e+00
|
||||
4.3543243243243239e+00
|
||||
4.3595121951219520e+00
|
||||
7.1674999999999995e+00
|
||||
6.0100000000000007e+00
|
||||
3.8176923076923064e+00
|
||||
3.8862962962962957e+00
|
||||
4.4436363636363652e+00
|
||||
4.8211111111111098e+00
|
||||
7.1261904761904749e+00
|
||||
3.7118181818181823e+00
|
||||
8.1500000000000004e+00
|
||||
7.7725000000000000e+00
|
||||
4.5788000000000011e+00
|
||||
4.1022222222222222e+00
|
||||
5.2209756097560973e+00
|
||||
5.7300000000000004e+00
|
||||
2.7129629629629628e+00
|
||||
2.4211111111111108e+00
|
||||
2.8095454545454532e+00
|
||||
2.4837209302325594e+00
|
||||
2.4050000000000002e+00
|
||||
4.2750000000000004e+00
|
||||
3.9422222222222221e+00
|
||||
3.8742307692307696e+00
|
||||
2.3256818181818173e+00
|
||||
2.6061538461538465e+00
|
||||
3.0470370370370370e+00
|
||||
1.8579545454545447e+00
|
||||
4.8434090909090912e+00
|
||||
4.7172093023255792e+00
|
||||
2.9300000000000002e+00
|
||||
3.2725000000000000e+00
|
||||
5.7482608695652173e+00
|
||||
4.5774074074074083e+00
|
||||
5.3236363636363633e+00
|
||||
5.9372499999999988e+00
|
||||
5.0633333333333335e+00
|
||||
7.6900000000000004e+00
|
||||
4.9976923076923079e+00
|
||||
5.0788888888888879e+00
|
||||
4.1337209302325579e+00
|
||||
4.6752272727272706e+00
|
||||
2.1450000000000000e+00
|
||||
2.9566666666666670e+00
|
||||
1.1678124999999998e+01
|
||||
1.0795000000000000e+01
|
||||
6.3430000000000009e+00
|
||||
5.9580952380952388e+00
|
||||
9.2199999999999989e+00
|
||||
6.3524999999999991e+00
|
||||
6.3756000000000004e+00
|
||||
7.0014285714285718e+00
|
||||
5.0088372093023246e+00
|
||||
6.2643243243243232e+00
|
||||
6.1133333333333333e+00
|
||||
5.3249999999999993e+00
|
||||
2.7474074074074073e+00
|
||||
4.6804545454545430e+00
|
||||
5.7633333333333336e+00
|
||||
5.0625000000000000e+00
|
||||
4.9824999999999999e+00
|
||||
6.3411999999999997e+00
|
||||
4.5371428571428583e+00
|
||||
3.5171428571428587e+00
|
||||
3.1532558139534879e+00
|
||||
5.2366666666666664e+00
|
||||
5.1399999999999988e+00
|
||||
8.4735714285714288e+00
|
||||
7.1446666666666658e+00
|
||||
3.5134090909090903e+00
|
||||
2.9665909090909079e+00
|
||||
6.0766666666666671e+00
|
||||
3.8666666666666667e+00
|
||||
4.9534615384615384e+00
|
||||
4.8718518518518508e+00
|
||||
4.8771428571428563e+00
|
||||
6.2160975609756104e+00
|
||||
6.8499999999999996e+00
|
||||
4.2533333333333330e+00
|
||||
4.9648148148148143e+00
|
||||
4.9369230769230761e+00
|
||||
5.1770833333333330e+00
|
||||
3.1988372093023254e+00
|
||||
3.6213636363636370e+00
|
||||
9.4099999999999984e+00
|
||||
7.7975000000000003e+00
|
||||
6.8578947368421055e+00
|
||||
7.5875000000000004e+00
|
||||
1.9634090909090893e+00
|
||||
2.4779545454545455e+00
|
||||
5.7725000000000000e+00
|
||||
5.6566666666666663e+00
|
||||
3.5634615384615382e+00
|
||||
3.4437037037037039e+00
|
||||
2.9709090909090894e+00
|
||||
2.8075000000000006e+00
|
||||
5.7266666666666666e+00
|
||||
5.0633333333333335e+00
|
||||
5.6592307692307688e+00
|
||||
4.8769230769230765e+00
|
||||
4.5544186046511639e+00
|
||||
5.0542105263157877e+00
|
||||
8.0066666666666677e+00
|
||||
6.9049999999999994e+00
|
||||
3.1661538461538474e+00
|
||||
2.7900000000000000e+00
|
||||
2.2986363636363647e+00
|
||||
2.8874418604651164e+00
|
||||
3.5874999999999999e+00
|
||||
4.4849999999999994e+00
|
||||
9.3166666666666682e+00
|
||||
9.7884615384615401e+00
|
||||
3.9845454545454548e+00
|
||||
3.8721052631578958e+00
|
||||
4.7133333333333338e+00
|
||||
3.8250000000000002e+00
|
||||
4.3996153846153847e+00
|
||||
3.8052000000000006e+00
|
||||
4.9980952380952370e+00
|
||||
4.2854545454545452e+00
|
||||
3.9000000000000004e+00
|
||||
3.5349999999999997e+00
|
||||
2.4480769230769228e+00
|
||||
3.1014814814814819e+00
|
||||
8.3771428571428572e+00
|
||||
9.8775000000000013e+00
|
||||
3.9299999999999997e+00
|
||||
5.6400000000000006e+00
|
||||
9.2333333333333343e+00
|
||||
7.3440909090909097e+00
|
||||
5.8252380952380962e+00
|
||||
6.1797674418604664e+00
|
||||
4.2533333333333330e+00
|
||||
4.2750000000000004e+00
|
||||
4.2941666666666665e+00
|
||||
4.6457692307692309e+00
|
||||
4.7833333333333332e+00
|
||||
3.3774999999999995e+00
|
||||
3.2266666666666675e+00
|
||||
2.3244444444444445e+00
|
||||
2.4297674418604669e+00
|
||||
2.5704761904761906e+00
|
||||
2.6066666666666669e+00
|
||||
2.1533333333333333e+00
|
||||
3.3025925925925925e+00
|
||||
3.0223809523809528e+00
|
||||
5.0890697674418597e+00
|
||||
4.8595348837209293e+00
|
||||
3.8475000000000001e+00
|
||||
4.0325000000000006e+00
|
||||
4.7977777777777764e+00
|
||||
4.2389285714285716e+00
|
||||
4.8243902439024371e+00
|
||||
6.7746666666666640e+00
|
||||
4.9933333333333332e+00
|
||||
1.2701333333333332e+01
|
||||
1.0377999999999998e+01
|
||||
9.7267999999999990e+00
|
||||
6.4125581395348812e+00
|
||||
3.2033333333333331e+00
|
||||
5.6233333333333340e+00
|
||||
2.8969999999999989e+00
|
||||
8.4913793103448274e+00
|
||||
5.1333333333333329e+00
|
||||
4.6433333333333335e+00
|
||||
2.9820000000000007e+00
|
||||
2.6970370370370373e+00
|
||||
2.9765909090909086e+00
|
||||
3.3188372093023251e+00
|
||||
4.4299999999999997e+00
|
||||
3.7174999999999998e+00
|
||||
4.2815384615384611e+00
|
||||
3.6144444444444441e+00
|
||||
3.9388095238095255e+00
|
||||
3.8513953488372095e+00
|
||||
4.3999999999999995e+00
|
||||
4.6066666666666665e+00
|
||||
4.5448148148148153e+00
|
||||
5.7560000000000002e+00
|
||||
5.3117857142857128e+00
|
||||
2.5165909090909095e+00
|
||||
2.8383720930232559e+00
|
||||
3.8250000000000002e+00
|
||||
6.2425000000000006e+00
|
||||
3.9725925925925933e+00
|
||||
3.7232142857142856e+00
|
||||
4.3255813953488396e+00
|
||||
4.5230952380952374e+00
|
||||
4.6066666666666665e+00
|
||||
4.2233333333333336e+00
|
||||
4.4750000000000014e+00
|
||||
4.5025925925925936e+00
|
||||
5.7267500000000009e+00
|
||||
6.3217647058823534e+00
|
||||
4.0433333333333339e+00
|
||||
4.2074999999999996e+00
|
||||
5.2319230769230769e+00
|
||||
5.7133333333333338e+00
|
||||
1.9861363636363627e+00
|
||||
2.6293181818181823e+00
|
||||
4.6733333333333329e+00
|
||||
6.7050000000000001e+00
|
||||
3.8875999999999999e+00
|
||||
5.0888000000000000e+00
|
||||
5.5153488372093014e+00
|
||||
7.4194444444444452e+00
|
||||
6.1099999999999994e+00
|
||||
5.3824999999999994e+00
|
||||
4.2187499999999991e+00
|
||||
3.8128571428571418e+00
|
||||
3.4400000000000004e+00
|
||||
4.1067441860465115e+00
|
||||
2.6999999999999997e+00
|
||||
2.8600000000000003e+00
|
||||
4.0196153846153839e+00
|
||||
3.8457692307692311e+00
|
||||
3.8817948717948707e+00
|
||||
3.7552272727272711e+00
|
||||
7.4275000000000002e+00
|
||||
8.6899999999999995e+00
|
||||
6.4753846153846162e+00
|
||||
6.0141666666666653e+00
|
||||
8.0438235294117675e+00
|
||||
5.9988235294117649e+00
|
||||
6.7933333333333339e+00
|
||||
3.1400000000000001e+00
|
||||
3.3200000000000003e+00
|
||||
4.1522222222222203e+00
|
||||
3.8577777777777786e+00
|
||||
3.3722727272727266e+00
|
||||
3.3793181818181810e+00
|
||||
4.2266666666666666e+00
|
||||
3.5549999999999997e+00
|
||||
4.5107407407407401e+00
|
||||
3.5775000000000001e+00
|
||||
6.9528571428571420e+00
|
||||
6.0760526315789498e+00
|
||||
4.3299999999999992e+00
|
||||
3.0000000000000000e+00
|
||||
7.0819999999999981e+00
|
||||
6.2977777777777781e+00
|
||||
3.2575000000000003e+00
|
||||
3.5443181818181810e+00
|
||||
9.6750000000000007e+00
|
||||
8.2336363636363643e+00
|
||||
8.7376923076923081e+00
|
||||
8.5203448275862055e+00
|
||||
7.6451724137931034e+00
|
||||
3.8666666666666667e+00
|
||||
3.5866666666666673e+00
|
||||
4.9269230769230763e+00
|
||||
5.9090909090909101e+00
|
||||
4.4325000000000001e+00
|
||||
5.4964102564102548e+00
|
||||
2.8775000000000004e+00
|
||||
3.2966666666666669e+00
|
||||
2.5385185185185191e+00
|
||||
2.6122222222222224e+00
|
||||
4.2142222222222214e+00
|
||||
4.1902631578947371e+00
|
||||
4.6266666666666660e+00
|
||||
3.8533333333333335e+00
|
||||
8.4165000000000010e+00
|
||||
5.5291666666666659e+00
|
||||
2.3225581395348835e+00
|
||||
3.0686363636363638e+00
|
||||
4.0800000000000001e+00
|
||||
3.7133333333333329e+00
|
||||
5.5911538461538459e+00
|
||||
6.5222727272727266e+00
|
||||
3.6168181818181817e+00
|
||||
3.4525000000000015e+00
|
||||
3.1800000000000002e+00
|
||||
4.3366666666666669e+00
|
||||
4.1024999999999991e+00
|
||||
6.4061111111111115e+00
|
||||
6.1661904761904767e+00
|
||||
8.1230000000000011e+00
|
||||
5.8008571428571418e+00
|
||||
6.4733333333333336e+00
|
||||
1.0175000000000001e+01
|
||||
5.0976190476190473e+00
|
||||
5.2099999999999991e+00
|
||||
3.5631818181818171e+00
|
||||
4.1541860465116276e+00
|
||||
5.4874999999999998e+00
|
||||
6.9100000000000001e+00
|
||||
9.7192857142857143e+00
|
||||
9.8164285714285722e+00
|
||||
5.6602564102564097e+00
|
||||
8.7630769230769214e+00
|
||||
1.0220000000000001e+01
|
||||
1.0695000000000000e+01
|
||||
3.3466666666666676e+00
|
||||
3.4067857142857143e+00
|
||||
3.1056818181818184e+00
|
||||
3.2893023255813940e+00
|
||||
5.6650000000000009e+00
|
||||
6.1125000000000007e+00
|
||||
9.8085714285714314e+00
|
||||
1.2693333333333333e+01
|
||||
6.8272222222222219e+00
|
||||
5.9281249999999996e+00
|
||||
4.7619999999999996e+00
|
||||
3.9049999999999998e+00
|
||||
6.0947500000000012e+00
|
||||
5.3353333333333337e+00
|
||||
4.9464705882352931e+00
|
||||
2.7939999999999992e+00
|
||||
3.6475000000000000e+00
|
||||
4.0997222222222245e+00
|
||||
4.8644444444444437e+00
|
||||
5.6488235294117644e+00
|
||||
7.6616666666666688e+00
|
||||
3.9233333333333325e+00
|
||||
4.2925806451612898e+00
|
||||
5.9369565217391322e+00
|
||||
5.0370588235294118e+00
|
||||
3.7346666666666666e+00
|
||||
4.6871428571428577e+00
|
||||
6.2183333333333337e+00
|
||||
4.9388888888888873e+00
|
||||
3.6206249999999986e+00
|
||||
7.6308823529411747e+00
|
||||
4.7541176470588242e+00
|
||||
4.6737209302325571e+00
|
||||
4.8757142857142863e+00
|
||||
6.9187179487179495e+00
|
||||
6.2716666666666656e+00
|
||||
3.2473333333333332e+00
|
||||
2.9200000000000004e+00
|
||||
2.6910810810810815e+00
|
||||
4.6559090909090921e+00
|
||||
8.3445454545454538e+00
|
||||
4.6804545454545439e+00
|
||||
2.2156249999999997e+00
|
||||
1.9752941176470586e+00
|
||||
2.5693478260869576e+00
|
||||
2.1659999999999995e+00
|
||||
5.6924999999999999e+00
|
||||
4.6219444444444466e+00
|
||||
3.3962222222222200e+00
|
||||
3.8658823529411754e+00
|
||||
5.9802439024390246e+00
|
||||
5.5507142857142862e+00
|
||||
7.6572727272727281e+00
|
||||
7.2237837837837819e+00
|
||||
4.6305555555555564e+00
|
||||
3.9100000000000006e+00
|
||||
4.6493750000000000e+00
|
||||
5.5844444444444434e+00
|
||||
3.0119565217391302e+00
|
||||
6.4043750000000008e+00
|
||||
9.5586206896551715e+00
|
||||
7.2057142857142855e+00
|
||||
5.6258333333333317e+00
|
||||
5.7640476190476182e+00
|
||||
7.1963636363636363e+00
|
||||
5.6874418604651167e+00
|
||||
5.4273333333333342e+00
|
||||
6.0615151515151524e+00
|
||||
5.5800000000000018e+00
|
||||
4.1017647058823528e+00
|
||||
2.6497777777777776e+00
|
||||
5.2453333333333330e+00
|
||||
7.3470967741935489e+00
|
||||
4.3131818181818193e+00
|
||||
5.5653333333333324e+00
|
||||
7.0499999999999980e+00
|
||||
3.7619999999999991e+00
|
||||
5.0675757575757592e+00
|
||||
5.9769999999999985e+00
|
||||
6.7249999999999988e+00
|
||||
8.6042307692307691e+00
|
||||
3.5806666666666662e+00
|
||||
3.5815625000000004e+00
|
||||
3.7788636363636381e+00
|
||||
3.7358333333333325e+00
|
||||
5.1269565217391317e+00
|
||||
4.8678571428571429e+00
|
||||
5.5264705882352940e+00
|
||||
4.2017073170731711e+00
|
||||
3.7105882352941166e+00
|
||||
5.3206818181818178e+00
|
||||
2.9713333333333343e+00
|
||||
2.7563888888888886e+00
|
||||
7.6529629629629641e+00
|
||||
6.4521428571428556e+00
|
||||
5.8050000000000006e+00
|
||||
2.0128571428571429e+00
|
||||
4.2888888888888896e+00
|
||||
5.9328571428571450e+00
|
||||
1.0609999999999999e+01
|
||||
3.7197777777777770e+00
|
||||
3.1337500000000009e+00
|
||||
3.2837837837837829e+00
|
||||
4.4520000000000000e+00
|
||||
4.6988235294117651e+00
|
||||
5.7946511627906983e+00
|
||||
4.8130769230769230e+00
|
||||
5.3478125000000007e+00
|
||||
6.2370731707317066e+00
|
||||
7.8106249999999990e+00
|
||||
3.8891304347826114e+00
|
||||
2.3959999999999999e+00
|
||||
4.2005405405405414e+00
|
||||
4.7605128205128180e+00
|
||||
6.9815384615384621e+00
|
||||
9.1547058823529408e+00
|
||||
4.0542857142857134e+00
|
||||
3.8008333333333320e+00
|
||||
3.1255813953488385e+00
|
||||
4.2582352941176467e+00
|
||||
4.8453488372093014e+00
|
||||
5.1150000000000002e+00
|
||||
4.7213513513513510e+00
|
||||
3.4827272727272724e+00
|
||||
3.0505882352941178e+00
|
||||
2.7824444444444429e+00
|
||||
7.7399999999999993e+00
|
||||
1.0409130434782606e+01
|
||||
5.9644444444444451e+00
|
||||
1.0795999999999999e+01
|
||||
8.9992592592592597e+00
|
||||
6.3976923076923091e+00
|
||||
7.1279310344827573e+00
|
||||
4.2302272727272721e+00
|
||||
3.2905882352941180e+00
|
||||
4.7038636363636357e+00
|
||||
3.4220000000000002e+00
|
||||
4.2258823529411762e+00
|
||||
3.2204444444444449e+00
|
||||
2.4581249999999999e+00
|
||||
2.1877777777777769e+00
|
||||
2.1423913043478255e+00
|
||||
2.2755555555555547e+00
|
||||
3.2992857142857148e+00
|
||||
4.5726666666666667e+00
|
||||
4.4771428571428586e+00
|
||||
4.4768181818181816e+00
|
||||
4.5383333333333340e+00
|
||||
4.2769565217391312e+00
|
||||
4.6015384615384614e+00
|
||||
5.7247058823529393e+00
|
||||
4.9804444444444425e+00
|
||||
3.6694444444444434e+00
|
||||
1.1215238095238094e+01
|
||||
9.1850000000000005e+00
|
||||
5.3722222222222227e+00
|
||||
6.2425000000000042e+00
|
||||
7.3538461538461544e+00
|
||||
3.0506250000000006e+00
|
||||
7.5833333333333321e+00
|
||||
4.0663636363636355e+00
|
||||
3.4388235294117653e+00
|
||||
3.5664444444444445e+00
|
||||
3.0406666666666671e+00
|
||||
3.8337142857142852e+00
|
||||
3.7924444444444458e+00
|
||||
5.3774999999999986e+00
|
||||
4.4091111111111108e+00
|
||||
4.1324999999999994e+00
|
||||
3.8061111111111114e+00
|
||||
4.5399999999999991e+00
|
||||
3.6055555555555543e+00
|
||||
3.9333333333333367e+00
|
||||
7.2324324324324314e+00
|
||||
2.5419999999999998e+00
|
||||
3.4575675675675668e+00
|
||||
5.8765116279069751e+00
|
||||
9.4742857142857133e+00
|
||||
3.7115217391304345e+00
|
||||
4.8933333333333326e+00
|
||||
6.4018181818181805e+00
|
||||
4.4633333333333329e+00
|
||||
3.6568750000000003e+00
|
||||
4.3422727272727268e+00
|
||||
4.2013333333333325e+00
|
||||
4.1880555555555556e+00
|
||||
3.4113636363636362e+00
|
||||
4.9605882352941180e+00
|
||||
4.7213333333333320e+00
|
||||
2.1199999999999992e+00
|
||||
3.2251351351351345e+00
|
||||
7.0423333333333336e+00
|
||||
4.8461111111111101e+00
|
||||
3.5000000000000000e+00
|
||||
4.7093333333333343e+00
|
||||
6.8564516129032258e+00
|
||||
5.3060465116279092e+00
|
||||
3.3566666666666656e+00
|
||||
4.4884444444444425e+00
|
||||
3.9066666666666667e+00
|
||||
4.9237837837837830e+00
|
||||
2.8633333333333337e+00
|
||||
2.3141176470588238e+00
|
||||
3.9447727272727278e+00
|
||||
3.9740000000000006e+00
|
||||
3.1935135135135138e+00
|
||||
7.9008333333333338e+00
|
||||
1.2878000000000000e+01
|
||||
7.3702380952380944e+00
|
||||
8.6866666666666656e+00
|
||||
5.4328571428571442e+00
|
||||
5.2213953488372109e+00
|
||||
9.0155555555555562e+00
|
||||
3.6631707317073179e+00
|
||||
3.6888888888888900e+00
|
||||
4.1529545454545467e+00
|
||||
3.5973333333333337e+00
|
||||
3.4672972972972973e+00
|
||||
3.3593333333333320e+00
|
||||
3.2805882352941182e+00
|
||||
4.6777777777777771e+00
|
||||
6.9690909090909088e+00
|
||||
5.7182758620689640e+00
|
||||
2.7732608695652186e+00
|
||||
3.6272222222222235e+00
|
||||
6.5997368421052620e+00
|
||||
3.5273333333333339e+00
|
||||
3.3930555555555548e+00
|
||||
1.0433571428571428e+01
|
||||
9.1080000000000005e+00
|
||||
8.6382608695652188e+00
|
||||
8.7341666666666669e+00
|
||||
7.3107692307692300e+00
|
||||
3.3476744186046501e+00
|
||||
2.6300000000000003e+00
|
||||
4.5486666666666657e+00
|
||||
3.9621428571428572e+00
|
||||
5.5491176470588242e+00
|
||||
2.5150000000000010e+00
|
||||
2.5094444444444446e+00
|
||||
2.1204651162790711e+00
|
||||
4.4307142857142860e+00
|
||||
4.5968571428571421e+00
|
||||
5.1136585365853655e+00
|
||||
4.6394444444444449e+00
|
||||
6.1210000000000004e+00
|
||||
2.0293750000000004e+00
|
||||
3.0340540540540539e+00
|
||||
3.5041304347826112e+00
|
||||
3.1037499999999993e+00
|
||||
5.1755813953488401e+00
|
||||
3.9831250000000002e+00
|
||||
4.4294594594594594e+00
|
||||
4.3844444444444450e+00
|
||||
3.9411111111111108e+00
|
||||
6.6910526315789474e+00
|
||||
8.6833333333333318e+00
|
||||
5.3197058823529408e+00
|
||||
6.9751612903225810e+00
|
||||
5.9539999999999997e+00
|
||||
4.9557777777777785e+00
|
||||
3.4481250000000006e+00
|
||||
4.9451428571428577e+00
|
||||
6.9590322580645161e+00
|
||||
8.6592307692307688e+00
|
||||
8.1816666666666666e+00
|
||||
5.0956249999999992e+00
|
||||
8.1888000000000005e+00
|
||||
1.0237058823529411e+01
|
||||
5.3599999999999994e+00
|
||||
3.6475555555555559e+00
|
||||
3.2793333333333341e+00
|
||||
5.1406060606060597e+00
|
||||
5.8326190476190503e+00
|
||||
5.1949999999999985e+00
|
||||
1.0530500000000000e+01
|
||||
6.6633333333333340e+00
|
||||
5.4306060606060624e+00
|
File diff suppressed because it is too large
Load Diff
File diff suppressed because it is too large
Load Diff
@ -1,485 +0,0 @@
|
||||
"""
|
||||
Data package in WAFO Toolbox.
|
||||
|
||||
Contents
|
||||
--------
|
||||
atlantic - Significant wave-height data recorded in the Atlantic Ocean
|
||||
gfaks89 - Surface elevation measured at Gullfaks C 24.12.1989
|
||||
gfaksr89 - Reconstructed surface elevation measured at Gullfaks C 24.12.1989.
|
||||
japansea - coastline map of The Japan Sea
|
||||
northsea - coastline map of The Nortsea
|
||||
sea - Surface elevation dataset used in WAT version 1.1.
|
||||
sfa89 - Wind measurements at Statfjord A 24.12.1989
|
||||
sn - Fatigue experiment, constant-amplitude loading.
|
||||
yura87 - Surface elevation measured off the coast of Yura
|
||||
|
||||
|
||||
|
||||
This module gives gives detailed information and easy access to all datasets
|
||||
included in WAFO
|
||||
|
||||
"""
|
||||
import numpy as np
|
||||
|
||||
import os
|
||||
__path2data = os.path.dirname(os.path.realpath(__file__))
|
||||
|
||||
__all__ = ['atlantic', 'gfaks89', 'gfaksr89', 'japansea', 'northsea', 'sea',
|
||||
'sfa89', 'sn', 'yura87']
|
||||
|
||||
_NANS = set(['nan', 'NaN', '-1.#IND00+00', '1.#IND00+00', '-1.#INF00+00'])
|
||||
|
||||
|
||||
def _tofloat(x):
|
||||
return np.nan if x in _NANS else float(x or 0)
|
||||
|
||||
|
||||
_MYCONVERTER = {}
|
||||
for i in range(2):
|
||||
_MYCONVERTER[i] = _tofloat
|
||||
|
||||
|
||||
def _load(file): # @ReservedAssignment
|
||||
""" local load function
|
||||
"""
|
||||
return np.loadtxt(os.path.join(__path2data, file))
|
||||
|
||||
|
||||
def _loadnan(file): # @ReservedAssignment
|
||||
""" local load function accepting nan's
|
||||
"""
|
||||
return np.loadtxt(os.path.join(__path2data, file), converters=_MYCONVERTER)
|
||||
|
||||
|
||||
def atlantic():
|
||||
"""
|
||||
Return Significant wave-height data recorded in the Atlantic Ocean
|
||||
|
||||
Data summary
|
||||
------------
|
||||
Size : 582 X 1
|
||||
Sampling Rate : ~ 14 times a month
|
||||
Device :
|
||||
Source :
|
||||
Format : ascii
|
||||
|
||||
Description
|
||||
------------
|
||||
atlantic.dat contains average significant wave-height data recorded
|
||||
approximately 14 times a month in December-February during 7 years and
|
||||
at 2 locations in the Atlantic Ocean
|
||||
|
||||
Example
|
||||
--------
|
||||
>>> import pylab
|
||||
>>> import wafo
|
||||
>>> Hs = wafo.data.atlantic()
|
||||
>>> np.allclose(Hs[:3], [ 5.48296296, 4.3615 , 5.26023256])
|
||||
True
|
||||
|
||||
h = pylab.plot(Hs)
|
||||
|
||||
Acknowledgement:
|
||||
---------------
|
||||
This dataset were made available by Dr. David Carter
|
||||
and Dr. David Cotton, Satellite Observing Systems, UK.
|
||||
"""
|
||||
return _load('atlantic.dat')
|
||||
|
||||
|
||||
def gfaks89():
|
||||
"""
|
||||
Return Surface elevation measured at Gullfaks C 24.12.1989
|
||||
|
||||
Data summary
|
||||
------------
|
||||
Size : 39000 X 2
|
||||
Sampling Rate : 2.5 Hz
|
||||
Device : EMI laser
|
||||
Source : STATOIL
|
||||
Format : ascii, c1: time c2: surface elevation
|
||||
|
||||
Description
|
||||
------------
|
||||
The wave data was measured 24th December 1989 at the Gullfaks C platform
|
||||
in the North Sea from 17.00 to 21.20. The period from 20.00 to 20.20
|
||||
is missing and contains NaNs. The water depth of 218 m is
|
||||
regarded as deep water for the most important wave components.
|
||||
There are two EMI laser sensors named 219 and 220. This data set is
|
||||
obtained from sensor 219, which is located in the Northwest
|
||||
corner approximately two platform leg diameters away from
|
||||
the closest leg.
|
||||
Thus the wave elevation is not expected to be significantly
|
||||
affected by diffraction effects for incoming waves in the western sector.
|
||||
The wind direction for this period is from the south.
|
||||
Some difficulties in calibration of the instruments have been reported
|
||||
resulting in several consecutive measured values being equal or almost
|
||||
equal in the observed data set.
|
||||
|
||||
This dataset is for non-commercial use only.
|
||||
|
||||
Hm0 = 6.8m, Tm02 = 8s, Tp = 10.5
|
||||
|
||||
Example
|
||||
-------
|
||||
>>> import pylab
|
||||
>>> import wafo
|
||||
>>> x = wafo.data.gfaks89()
|
||||
>>> np.allclose(x[:3, 1], [-0.19667949, -0.46667949, -0.38667949])
|
||||
True
|
||||
|
||||
h = pylab.plot(x[:,0],x[:,1])
|
||||
|
||||
Acknowledgement:
|
||||
---------------
|
||||
This dataset were prepared and made available by Dr. S. Haver,
|
||||
STATOIL, Norway
|
||||
|
||||
See also
|
||||
--------
|
||||
gfaksr89, northsea
|
||||
|
||||
"""
|
||||
return _loadnan('gfaks89.dat')
|
||||
|
||||
|
||||
def gfaksr89():
|
||||
"""
|
||||
Return a reconstruction of surface elevation measured at Gullfaks C
|
||||
24.12.1989.
|
||||
|
||||
|
||||
Data summary
|
||||
------------
|
||||
Size : 39000 X 2
|
||||
Sampling Rate : 2.5 Hz
|
||||
Device : EMI laser
|
||||
Source : STATOIL
|
||||
Format : ascii, c1: time c2: surface elevation
|
||||
|
||||
Description
|
||||
-----------
|
||||
This is a reconstructed version of the data in the GFAKS89.DAT file.
|
||||
The following calls were made to reconstruct the data:
|
||||
|
||||
inds = findoutliers(gfaks89,.02,2,1.23);
|
||||
gfaksr89 = reconstruct(gfaks89,inds,6);
|
||||
|
||||
The wave data was measured 24th December 1989 at the Gullfaks C platform
|
||||
in the North Sea from 17.00 to 21.20. The period from 20.00 to 20.20
|
||||
is missing in the original data. The water depth of 218 m is
|
||||
regarded as deep water for the most important wave components.
|
||||
There are two EMI laser sensors named 219 and 220. This data set is
|
||||
obtained from sensor 219, which is located in the Northwest
|
||||
corner approximately two platform leg diameters away from
|
||||
the closest leg.
|
||||
Thus the wave elevation is not expected to be significantly
|
||||
affected by diffraction effects for incoming waves in the western sector.
|
||||
The wind direction for this period is from the south.
|
||||
Some difficulties in calibration of the instruments have been reported
|
||||
resulting in several consecutive measured values being equal or almost
|
||||
equal in the observed data set.
|
||||
|
||||
Hm0 = 6.8m, Tm02 = 8s, Tp = 10.5
|
||||
|
||||
|
||||
Example
|
||||
-------
|
||||
>>> import pylab
|
||||
>>> import wafo
|
||||
>>> x = wafo.data.gfaksr89()
|
||||
|
||||
h = pylab.plot(x[:,0],x[:,1])
|
||||
|
||||
See also
|
||||
--------
|
||||
gfaks89
|
||||
"""
|
||||
return _loadnan('gfaksr89.dat')
|
||||
|
||||
|
||||
def japansea():
|
||||
"""
|
||||
Return coastline map of The Japan Sea
|
||||
|
||||
|
||||
Data summary
|
||||
------------
|
||||
Size : 692 X 2
|
||||
Sampling Rate :
|
||||
Device :
|
||||
Source : http://crusty.er.usgs.gov/coast/getcoast.html
|
||||
Format : ascii, c1: longitude c2: latitude
|
||||
|
||||
Description
|
||||
-----------
|
||||
JAPANSEA.DAT contains data for plotting a map of The Japan Sea.
|
||||
The data is obtained from USGS coastline extractor.
|
||||
|
||||
Example:
|
||||
-------
|
||||
#the map is seen by
|
||||
|
||||
>>> import pylab
|
||||
>>> import wafo
|
||||
>>> map1 = wafo.data.japansea()
|
||||
>>> np.allclose(map1[1:4, 0], [ 141.960057, 142.058624, 142.103214])
|
||||
True
|
||||
|
||||
h = pylab.plot(map1[:,0],map1[:,1])
|
||||
lon_loc = [131,132,132,135,139.5,139]
|
||||
lat_loc = [46, 43, 40, 35, 38.3, 35.7]
|
||||
loc = ['China','Vladivostok','Japan Sea', 'Japan', 'Yura','Tokyo']
|
||||
algn = 'right'
|
||||
for lon, lat, name in zip(lon_loc,lat_loc,loc):
|
||||
pylab.text(lon,lat,name,horizontalalignment=algn)
|
||||
|
||||
|
||||
# If you have the m_map toolbox (see http://www.ocgy.ubc.ca/~rich/):
|
||||
m_proj('lambert','long',[130 148],'lat',[30 48]);
|
||||
m_line(map(:,1),map(:,2));
|
||||
m_grid('box','fancy','tickdir','out');
|
||||
m_text(131,46,'China');
|
||||
m_text(132,43,'Vladivostok');
|
||||
m_text(132,40,'Japan Sea');
|
||||
m_text(135,35,'Japan');
|
||||
m_text(139.5,38.3,'Yura');
|
||||
m_text(139,35.7,'Tokyo');
|
||||
"""
|
||||
return _loadnan('japansea.dat')
|
||||
|
||||
|
||||
def northsea():
|
||||
"""
|
||||
NORTHSEA coastline map of The Nortsea
|
||||
|
||||
Data summary
|
||||
-------------
|
||||
Size : 60646 X 2
|
||||
Sampling Rate :
|
||||
Device :
|
||||
Source : http://crusty.er.usgs.gov/coast/getcoast.html
|
||||
Format : ascii, c1: longitude c2: latitude
|
||||
|
||||
Description
|
||||
-----------
|
||||
NORTHSEA.DAT contains data for plotting a map of The Northsea.
|
||||
The data is obtained from USGS coastline extractor.
|
||||
|
||||
Example
|
||||
-------
|
||||
# the map is seen by
|
||||
|
||||
>>> import pylab
|
||||
>>> import wafo
|
||||
>>> map1 = wafo.data.northsea()
|
||||
>>> np.allclose(map1[1:4, 0], [ 1.261996, 1.264064, 1.268171])
|
||||
True
|
||||
|
||||
h = pylab.plot(map1[:,0],map1[:,1])
|
||||
lon_pltfrm = [1.8, 2.3, 2., 1.9, 2.6]
|
||||
lat_pltfrm = [61.2, 61.2, 59.9, 58.4, 57.7]
|
||||
pltfrm = ['Statfjord A', 'Gullfaks C', 'Frigg', 'Sleipner', 'Draupner']
|
||||
h = pylab.scatter(lon_pltfrm,lat_pltfrm);
|
||||
algn = 'right'
|
||||
for lon, lat, name in zip(lon_pltfrm,lat_pltfrm,pltfrm):
|
||||
pylab.text(lon,lat,name,horizontalalignment=algn); algn = 'left'
|
||||
|
||||
|
||||
lon_city = [10.8, 10.8, 5.52, 5.2]
|
||||
lat_city = [59.85, 63.4, 58.9, 60.3]
|
||||
city = ['Oslo','Trondheim','Stavanger', 'Bergen']
|
||||
h = pylab.scatter(lon_city,lat_city);
|
||||
algn = 'right'
|
||||
for lon, lat, name in zip(lon_city,lat_city,city):
|
||||
pylab.text(lon,lat,name,horizontalalignment=algn)
|
||||
|
||||
# If you have the mpl_toolkits.basemap installed
|
||||
from mpl_toolkits.basemap import Basemap
|
||||
import matplotlib.pyplot as plt
|
||||
|
||||
# setup Lambert Conformal basemap.
|
||||
m = Basemap(width=1200000,height=900000,projection='lcc',
|
||||
resolution='f',lat_1=56.,lat_2=64,lat_0=58,lon_0=5.)
|
||||
m.drawcoastlines()
|
||||
h = m.scatter(lon_pltfrm,lat_pltfrm);
|
||||
algn = 'right'
|
||||
for lon, lat, name in zip(lon_pltfrm,lat_pltfrm,pltfrm):
|
||||
m.text(lon,lat,name,horizontalalignment=algn); algn = 'left'
|
||||
m.scatter(lon_city,lat_city)
|
||||
algn = 'right'
|
||||
for lon, lat, name in zip(lon_city,lat_city,city):
|
||||
m.text(lon,lat,name,horizontalalignment=algn)
|
||||
"""
|
||||
return _loadnan('northsea.dat')
|
||||
|
||||
|
||||
def sea():
|
||||
"""
|
||||
Return Surface elevation dataset used in WAT version 1.1.
|
||||
|
||||
Data summary
|
||||
------------
|
||||
Size : 9524 X 2
|
||||
Sampling Rate : 4.0 Hz
|
||||
Device : unknown
|
||||
Source : unknown
|
||||
Format : ascii, c1: time c2: surface elevation
|
||||
|
||||
Description
|
||||
-----------
|
||||
The wave data was used in one of WAFO predecessors, i.e. the Wave
|
||||
Analysis Toolbox version 1.1 (WAT)
|
||||
Hm0 = 1.9m, Tm02 = 4.0s, Tp2 = 11.5s Tp1=5.6s
|
||||
|
||||
Example
|
||||
-------
|
||||
>>> import pylab
|
||||
>>> import wafo
|
||||
>>> x = wafo.data.sea()
|
||||
>>> np.allclose(x[:3,1],[-1.2004945 , -1.0904945 , -0.79049454])
|
||||
True
|
||||
|
||||
h = pylab.plot(x[:,0],x[:,1])
|
||||
"""
|
||||
return _load('sea.dat')
|
||||
|
||||
|
||||
def sfa89():
|
||||
"""
|
||||
Return Wind measurements at Statfjord A 24.12.1989
|
||||
|
||||
Data summary
|
||||
------------
|
||||
Size : 144 X 3
|
||||
Sampling Rate : 1/600 Hz
|
||||
Device :
|
||||
Source : DNMI (The Norwegian Meteorological Institute)
|
||||
Format : ascii, c1: time (hours)
|
||||
c2: velocity (m/s)
|
||||
c3: direction (degrees)
|
||||
Description
|
||||
-----------
|
||||
The registration of wind speeds at the Gullfaks field
|
||||
started up on Statfjord A in 1978 and continued until 1990.
|
||||
The dataregistration was transferred to Gullfaks C in Nov 1989.
|
||||
Due to some difficulties of the windregistration on Gullfaks C in
|
||||
the beginning, they continued to use the registered data from
|
||||
Statfjord A.
|
||||
The windspeed is measured in (meter/second), 110 m above mean water
|
||||
level (MWL) and the wind direction is given in degrees for the data.
|
||||
The data are a mean value of every 10 minutes.
|
||||
Wind directions are defined in the meteorological convention, i.e.,
|
||||
0 degrees = wind approaching from North, 90 degrees = wind from East, etc.
|
||||
This dataset is for non-commercial use only.
|
||||
|
||||
Example
|
||||
-------
|
||||
>>> import pylab
|
||||
>>> import wafo
|
||||
>>> x = wafo.data.sfa89()
|
||||
|
||||
h = pylab.plot(x[:,0],x[:,1])
|
||||
|
||||
Acknowledgement
|
||||
----------------
|
||||
These data are made available by Knut A. Iden, DNMI.
|
||||
|
||||
See also
|
||||
--------
|
||||
northsea
|
||||
"""
|
||||
return _load('sfa89.dat')
|
||||
|
||||
|
||||
def sn():
|
||||
"""
|
||||
Return SN Fatigue experiment, constant-amplitude loading.
|
||||
|
||||
|
||||
Data summary
|
||||
------------
|
||||
Size : 40 X 2
|
||||
Source : unknown
|
||||
Format : ascii, c1: Amplitude MPa c2: Number of cycles
|
||||
|
||||
Description
|
||||
-----------
|
||||
A fatigue experiment with constant amplitudes at five levels:
|
||||
10,15,20,25 and 30 MPa. For each level is related 8 observations of
|
||||
the number of cycles to failure.
|
||||
|
||||
The origin of the data is unknown.
|
||||
|
||||
Example
|
||||
-------
|
||||
>>> import pylab
|
||||
>>> import wafo
|
||||
>>> x = wafo.data.sn()
|
||||
|
||||
h = pylab.plot(x[:,0],x[:,1])
|
||||
|
||||
See also
|
||||
--------
|
||||
The same data appear in the directory wdemos/itmkurs/
|
||||
as SN.mat.
|
||||
|
||||
"""
|
||||
return _load('sn.dat')
|
||||
|
||||
|
||||
def yura87():
|
||||
"""
|
||||
Return Surface elevation measured off the coast of Yura.
|
||||
|
||||
|
||||
Data summary
|
||||
-----------
|
||||
Size : 85547 X 4
|
||||
Sampling Rate : 1 Hz
|
||||
Device : ultrasonic wave gauges
|
||||
Source : SRI, Ministry of Transport, Japan
|
||||
Format : ascii, c1: time (sec) c2-4: surface elevation (m)
|
||||
|
||||
Description
|
||||
-----------
|
||||
The wave data was measured at the Poseidon platform
|
||||
in the Japan Sea from 24th November 1987 08.12 hours to 25th November
|
||||
1987 07.57 hours. Poseidon was located 3 km off the coast of Yura
|
||||
in the Yamagata prefecture, in the Japan Sea during the measurements.
|
||||
The most important wave components are to some extent influenced by the
|
||||
water depth of 42 m. The data are measured with three ultrasonic wave
|
||||
gauges located at the sea floor and the relative coordinates of the
|
||||
gauges are as follows (x-axis points to the East, y-axis points to
|
||||
the North):
|
||||
X (m) Y (m)
|
||||
c2: -4.93, 25.02
|
||||
c3: 5.80, 92.12
|
||||
c4: 0.00, 0.00
|
||||
|
||||
This dataset is for non-commercial use only.
|
||||
|
||||
Hm0 = 5.1m, Tm02 = 7.7s, Tp = 12.8s
|
||||
Example
|
||||
-------
|
||||
>>> import pylab
|
||||
>>> import wafo
|
||||
>>> x = wafo.data.yura87()
|
||||
|
||||
h = pylab.plot(x[:,0],x[:,1])
|
||||
|
||||
Acknowledgement:
|
||||
-----------------
|
||||
This dataset were prepared and made available by Dr. Sc. H. Tomita,
|
||||
Ship Research Institute, Ministry of Transport, Japan.
|
||||
|
||||
See also
|
||||
--------
|
||||
japansea
|
||||
"""
|
||||
return _load('yura87.dat')
|
||||
|
||||
|
||||
if __name__ == '__main__':
|
||||
import doctest
|
||||
doctest.testmod()
|
@ -1,692 +0,0 @@
|
||||
nan nan
|
||||
141.960057 40.022926
|
||||
142.058624 39.818752
|
||||
142.103214 39.640392
|
||||
142.157191 39.469073
|
||||
142.136070 39.332957
|
||||
142.070358 39.229696
|
||||
142.046890 39.074805
|
||||
141.927201 38.955116
|
||||
141.805166 38.948076
|
||||
141.727720 38.765023
|
||||
141.638541 38.542073
|
||||
141.638541 38.368408
|
||||
141.629153 38.316777
|
||||
141.396817 38.333205
|
||||
141.143358 38.194742
|
||||
141.054179 37.934243
|
||||
141.087034 37.671397
|
||||
141.143358 37.453142
|
||||
141.143358 37.143360
|
||||
141.164480 36.957960
|
||||
140.976733 36.826537
|
||||
140.854698 36.603588
|
||||
140.812455 36.533182
|
||||
140.756131 36.258603
|
||||
140.744396 35.944127
|
||||
140.887553 35.791583
|
||||
140.922756 35.683628
|
||||
140.767865 35.603836
|
||||
140.589505 35.359765
|
||||
140.479204 35.162631
|
||||
140.237480 35.007740
|
||||
nan nan
|
||||
139.984022 35.233036
|
||||
140.016877 35.369152
|
||||
140.148300 35.495881
|
||||
140.192890 35.622610
|
||||
nan nan
|
||||
139.972288 39.161638
|
||||
140.070855 39.323569
|
||||
140.138913 39.530091
|
||||
140.148300 39.750693
|
||||
140.016877 39.844567
|
||||
nan nan
|
||||
140.237480 35.007740
|
||||
139.993409 34.862236
|
||||
139.951166 35.007740
|
||||
139.984022 35.233036
|
||||
nan nan
|
||||
140.192890 35.622610
|
||||
139.951166 35.622610
|
||||
139.894842 35.451291
|
||||
139.805662 35.305788
|
||||
139.784541 35.153243
|
||||
139.653118 35.233036
|
||||
139.409047 35.261198
|
||||
139.298746 35.115694
|
||||
139.254156 34.918560
|
||||
139.209566 34.726119
|
||||
139.089878 34.618165
|
||||
138.956108 34.636940
|
||||
138.911518 34.871623
|
||||
138.946721 35.035902
|
||||
138.791830 35.052329
|
||||
138.625204 34.925600
|
||||
138.460926 34.763669
|
||||
138.348278 34.580616
|
||||
138.315422 34.554800
|
||||
138.195733 34.571228
|
||||
137.996252 34.608778
|
||||
137.731060 34.618165
|
||||
137.477602 34.618165
|
||||
137.256999 34.554800
|
||||
137.212409 34.554800
|
||||
137.278121 34.636940
|
||||
137.388422 34.709692
|
||||
137.266387 34.726119
|
||||
137.167820 34.744894
|
||||
137.113842 34.798871
|
||||
137.057518 34.709692
|
||||
136.935483 34.817646
|
||||
136.926096 35.017127
|
||||
136.836916 34.981924
|
||||
136.714880 34.754281
|
||||
136.628048 34.618165
|
||||
136.771204 34.536026
|
||||
136.968339 34.416337
|
||||
136.968339 34.179307
|
||||
136.836916 34.270833
|
||||
136.714880 34.216856
|
||||
136.583458 34.198081
|
||||
136.449688 34.134717
|
||||
136.372243 34.040844
|
||||
136.351121 33.977479
|
||||
136.229086 33.813201
|
||||
136.064807 33.555049
|
||||
135.853592 33.454135
|
||||
135.701048 33.498725
|
||||
135.579012 33.555049
|
||||
135.478098 33.702899
|
||||
135.290352 33.803813
|
||||
135.236374 33.977479
|
||||
135.248108 34.134717
|
||||
135.248108 34.261446
|
||||
135.334941 34.334198
|
||||
135.501567 34.564188
|
||||
135.522688 34.636940
|
||||
135.478098 34.662755
|
||||
135.325554 34.643980
|
||||
135.147195 34.608778
|
||||
134.926592 34.681530
|
||||
134.684868 34.726119
|
||||
134.464266 34.716732
|
||||
134.264785 34.580616
|
||||
134.142749 34.526638
|
||||
133.978471 34.435112
|
||||
133.955002 34.425725
|
||||
133.901025 34.463274
|
||||
133.835314 34.489089
|
||||
133.680423 34.444499
|
||||
133.579509 34.435112
|
||||
133.403496 34.371747
|
||||
133.281461 34.362360
|
||||
133.161772 34.308383
|
||||
133.016268 34.280221
|
||||
132.851990 34.233284
|
||||
132.664243 34.198081
|
||||
132.530473 34.298995
|
||||
132.420172 34.270833
|
||||
132.300484 34.015028
|
||||
132.244160 33.885952
|
||||
132.154980 33.867178
|
||||
131.924990 33.932889
|
||||
131.758365 34.005641
|
||||
131.570618 33.949317
|
||||
131.427461 33.949317
|
||||
131.338281 33.914114
|
||||
131.162269 33.968092
|
||||
130.986256 33.996254
|
||||
130.974522 34.207469
|
||||
130.997990 34.317770
|
||||
131.195124 34.362360
|
||||
131.460317 34.416337
|
||||
131.636329 34.571228
|
||||
131.814689 34.681530
|
||||
132.122124 34.827033
|
||||
132.321605 35.045289
|
||||
132.563329 35.179059
|
||||
132.718220 35.387927
|
||||
132.917701 35.486494
|
||||
133.138304 35.549859
|
||||
133.304929 35.521697
|
||||
133.459820 35.477107
|
||||
133.800111 35.505269
|
||||
134.077038 35.495881
|
||||
134.309374 35.540471
|
||||
134.583954 35.594448
|
||||
134.795169 35.631998
|
||||
134.959448 35.639038
|
||||
135.137807 35.693015
|
||||
135.325554 35.702403
|
||||
135.313820 35.559246
|
||||
135.412387 35.531084
|
||||
135.733903 35.477107
|
||||
135.820736 35.495881
|
||||
135.909916 35.568633
|
||||
136.031951 35.631998
|
||||
136.142253 35.693015
|
||||
136.142253 35.944127
|
||||
136.151640 36.176464
|
||||
136.440301 36.382985
|
||||
136.682025 36.631750
|
||||
136.836916 36.932144
|
||||
136.848650 37.143360
|
||||
136.893240 37.312332
|
||||
137.233531 37.469570
|
||||
137.388422 37.453142
|
||||
137.355566 37.312332
|
||||
137.212409 37.248967
|
||||
137.057518 37.126932
|
||||
137.036397 37.082342
|
||||
137.113842 36.915717
|
||||
137.113842 36.737357
|
||||
137.224144 36.727970
|
||||
137.433012 36.756132
|
||||
137.609025 36.906329
|
||||
137.853095 37.011937
|
||||
138.151144 37.091729
|
||||
138.470313 37.284170
|
||||
138.681528 37.453142
|
||||
138.836419 37.654970
|
||||
139.012432 37.847410
|
||||
139.233035 37.960058
|
||||
139.244769 37.969446
|
||||
139.254156 37.976486
|
||||
139.355070 37.976486
|
||||
139.507614 38.152499
|
||||
139.531083 38.307390
|
||||
139.575672 38.438813
|
||||
139.695361 38.628906
|
||||
139.850252 38.835428
|
||||
139.927698 38.997359
|
||||
139.972288 39.161638
|
||||
nan nan
|
||||
140.016877 39.844567
|
||||
139.805662 39.912625
|
||||
139.906576 39.980683
|
||||
nan nan
|
||||
138.604083 38.272187
|
||||
138.580614 38.098521
|
||||
138.625204 38.056278
|
||||
138.514903 37.812208
|
||||
138.360012 37.802820
|
||||
138.327156 37.950671
|
||||
138.449192 38.204129
|
||||
138.604083 38.272187
|
||||
nan nan
|
||||
134.276519 33.268735
|
||||
134.065304 33.379036
|
||||
133.800111 33.454135
|
||||
133.570121 33.397811
|
||||
133.391762 33.296897
|
||||
133.293195 33.111497
|
||||
133.105448 32.879160
|
||||
133.084326 32.712535
|
||||
132.905967 32.693760
|
||||
132.751076 32.804061
|
||||
132.685365 32.879160
|
||||
132.553942 33.083335
|
||||
132.575063 33.167821
|
||||
132.509352 33.278122
|
||||
132.387316 33.360261
|
||||
132.223038 33.360261
|
||||
132.488230 33.536274
|
||||
132.697099 33.756876
|
||||
132.896580 33.977479
|
||||
133.105448 33.949317
|
||||
133.239218 33.895340
|
||||
133.436352 33.932889
|
||||
133.624099 34.005641
|
||||
133.689810 34.134717
|
||||
133.823580 34.226244
|
||||
133.933881 34.317770
|
||||
134.065304 34.334198
|
||||
134.264785 34.289608
|
||||
134.452531 34.198081
|
||||
134.619157 34.188694
|
||||
134.661400 33.996254
|
||||
134.717724 33.822588
|
||||
134.762314 33.803813
|
||||
134.771701 33.794426
|
||||
134.762314 33.766264
|
||||
134.652012 33.665350
|
||||
134.551098 33.618413
|
||||
134.452531 33.517499
|
||||
134.363352 33.369649
|
||||
134.276519 33.268735
|
||||
nan nan
|
||||
135.060362 34.571228
|
||||
135.027506 34.526638
|
||||
134.872615 34.371747
|
||||
134.762314 34.207469
|
||||
134.872615 34.179307
|
||||
134.982916 34.226244
|
||||
135.004038 34.390522
|
||||
135.060362 34.571228
|
||||
nan nan
|
||||
130.864220 33.876565
|
||||
130.920544 33.885952
|
||||
130.831365 33.895340
|
||||
130.676474 33.848403
|
||||
130.521583 33.721674
|
||||
130.423015 33.564436
|
||||
130.279859 33.564436
|
||||
130.178945 33.470563
|
||||
130.035788 33.416585
|
||||
nan nan
|
||||
130.000000 32.676758
|
||||
130.146089 32.712535
|
||||
130.324448 32.628049
|
||||
130.411281 32.759472
|
||||
130.289246 32.813449
|
||||
130.235269 32.897935
|
||||
130.235269 33.083335
|
||||
130.333836 33.149046
|
||||
130.444137 32.944872
|
||||
130.577907 32.766512
|
||||
130.577907 32.599887
|
||||
130.622496 32.581112
|
||||
130.631884 32.581112
|
||||
130.622496 32.386325
|
||||
130.521583 32.179803
|
||||
130.401894 32.095317
|
||||
130.300980 31.935732
|
||||
130.289246 31.710436
|
||||
130.345570 31.464019
|
||||
130.289246 31.323208
|
||||
130.401894 31.229335
|
||||
130.655352 31.123727
|
||||
130.709329 31.341983
|
||||
130.631884 31.520343
|
||||
130.688208 31.663500
|
||||
130.852486 31.625950
|
||||
130.775041 31.569626
|
||||
130.753919 31.445244
|
||||
130.810243 31.248110
|
||||
130.810243 31.011079
|
||||
130.941666 31.048629
|
||||
131.096557 31.161277
|
||||
131.108291 31.304434
|
||||
131.251448 31.435857
|
||||
131.361750 31.388920
|
||||
131.483785 31.625950
|
||||
131.549496 31.879408
|
||||
131.615208 32.114092
|
||||
131.725509 32.386325
|
||||
131.835810 32.552950
|
||||
131.967233 32.712535
|
||||
132.023557 32.794674
|
||||
132.035291 32.841611
|
||||
132.023557 32.998849
|
||||
131.978967 33.073947
|
||||
131.924990 33.174861
|
||||
131.680919 33.221798
|
||||
131.648063 33.306284
|
||||
131.746630 33.517499
|
||||
131.692653 33.627801
|
||||
131.537762 33.583211
|
||||
131.239714 33.573823
|
||||
131.051967 33.766264
|
||||
131.040233 33.857790
|
||||
130.962788 33.857790
|
||||
130.864220 33.876565
|
||||
nan nan
|
||||
130.146089 32.477851
|
||||
130.092112 32.358163
|
||||
130.080378 32.217353
|
||||
130.113233 32.142254
|
||||
130.247003 32.421527
|
||||
130.146089 32.477851
|
||||
nan nan
|
||||
131.117679 30.630892
|
||||
131.084823 30.764662
|
||||
131.030846 30.630892
|
||||
130.962788 30.410290
|
||||
131.030846 30.344578
|
||||
131.117679 30.630892
|
||||
nan nan
|
||||
130.545051 30.353966
|
||||
130.512195 30.410290
|
||||
130.455871 30.325804
|
||||
130.512195 30.210809
|
||||
130.676474 30.192034
|
||||
130.676474 30.325804
|
||||
130.545051 30.353966
|
||||
nan nan
|
||||
130.897076 37.539975
|
||||
130.843099 37.495385
|
||||
130.843099 37.478957
|
||||
130.885342 37.453142
|
||||
130.974522 37.504772
|
||||
130.897076 37.539975
|
||||
nan nan
|
||||
130.035788 33.416585
|
||||
130.000000 33.445983
|
||||
nan nan
|
||||
142.678694 48.000000
|
||||
142.654720 47.880130
|
||||
142.666454 47.650140
|
||||
142.809611 47.448313
|
||||
142.952768 47.335665
|
||||
143.107659 47.194854
|
||||
143.140515 47.028229
|
||||
143.206226 46.915581
|
||||
143.307140 46.840482
|
||||
143.462031 46.772424
|
||||
143.516009 46.802933
|
||||
143.572333 46.755996
|
||||
143.626310 46.582330
|
||||
143.670900 46.415705
|
||||
143.649778 46.225612
|
||||
143.548864 46.070720
|
||||
143.494887 46.270201
|
||||
143.417442 46.537741
|
||||
143.173371 46.605799
|
||||
142.943381 46.687938
|
||||
142.699310 46.697325
|
||||
142.523297 46.558862
|
||||
142.434118 46.361728
|
||||
142.356672 46.148166
|
||||
142.258105 45.955725
|
||||
142.025768 46.049599
|
||||
141.969444 46.354687
|
||||
141.960057 46.657429
|
||||
142.025768 46.983639
|
||||
142.070358 47.194854
|
||||
142.091480 47.441272
|
||||
142.124335 47.671262
|
||||
142.157191 47.894211
|
||||
142.248875 48.000000
|
||||
nan nan
|
||||
148.000000 44.948933
|
||||
147.909284 44.948933
|
||||
147.789595 44.885569
|
||||
147.679294 44.754146
|
||||
147.566646 44.697822
|
||||
147.435223 44.620376
|
||||
147.313188 44.524156
|
||||
147.214621 44.446711
|
||||
147.071464 44.446711
|
||||
147.125441 44.540584
|
||||
147.235742 44.650885
|
||||
147.280332 44.768227
|
||||
147.381246 44.801083
|
||||
147.590114 45.002910
|
||||
147.733271 45.113212
|
||||
147.864694 45.160148
|
||||
147.953874 45.291571
|
||||
148.000000 45.306694
|
||||
nan nan
|
||||
146.937694 43.843574
|
||||
147.015140 43.820106
|
||||
146.928307 43.756741
|
||||
146.761682 43.691030
|
||||
146.663115 43.730926
|
||||
146.761682 43.796637
|
||||
146.937694 43.843574
|
||||
nan nan
|
||||
146.210175 44.446711
|
||||
146.231297 44.486607
|
||||
146.287621 44.446711
|
||||
146.365066 44.399774
|
||||
146.419044 44.303554
|
||||
146.320477 44.249576
|
||||
146.120996 44.097032
|
||||
145.912127 43.939794
|
||||
145.724380 43.796637
|
||||
145.614079 43.667561
|
||||
145.536634 43.796637
|
||||
145.703259 43.946835
|
||||
145.900393 44.113460
|
||||
146.067018 44.280085
|
||||
146.144464 44.430283
|
||||
146.210175 44.446711
|
||||
nan nan
|
||||
145.083695 44.066523
|
||||
145.149406 44.113460
|
||||
145.370008 44.256617
|
||||
145.447454 44.209680
|
||||
145.348887 44.003159
|
||||
145.182262 43.780209
|
||||
145.250320 43.620625
|
||||
145.426332 43.540832
|
||||
145.426332 43.339004
|
||||
145.646935 43.331964
|
||||
145.912127 43.404716
|
||||
145.813560 43.249825
|
||||
145.492044 43.146564
|
||||
145.226851 43.001060
|
||||
145.050839 42.977592
|
||||
144.952272 43.017488
|
||||
144.841970 42.904840
|
||||
144.508720 42.935349
|
||||
144.234140 42.911880
|
||||
143.957214 42.799232
|
||||
143.692021 42.538734
|
||||
143.516009 42.285275
|
||||
143.450297 42.048245
|
||||
143.328262 41.923863
|
||||
143.051335 42.081101
|
||||
142.842467 42.196096
|
||||
142.478708 42.294663
|
||||
142.180659 42.440166
|
||||
141.960057 42.529346
|
||||
141.650275 42.555161
|
||||
141.417938 42.473022
|
||||
141.253660 42.367415
|
||||
141.054179 42.383842
|
||||
140.845310 42.505878
|
||||
140.556650 42.489450
|
||||
140.404105 42.311091
|
||||
140.502672 42.153853
|
||||
140.702153 42.097529
|
||||
140.922756 42.015389
|
||||
141.110503 41.900394
|
||||
141.241925 41.808868
|
||||
141.176214 41.710301
|
||||
140.955612 41.726729
|
||||
140.788986 41.719688
|
||||
140.624708 41.635202
|
||||
140.523794 41.520207
|
||||
140.270336 41.388784
|
||||
140.138913 41.421640
|
||||
140.082589 41.562450
|
||||
140.160034 41.801827
|
||||
140.216358 41.982534
|
||||
140.028612 42.120997
|
||||
nan nan
|
||||
139.927698 42.522306
|
||||
140.094323 42.660769
|
||||
140.326660 42.749949
|
||||
140.514407 42.904840
|
||||
140.568384 43.057384
|
||||
140.469817 43.162992
|
||||
140.514407 43.282680
|
||||
140.744396 43.242784
|
||||
140.964999 43.179420
|
||||
141.152746 43.179420
|
||||
141.363961 43.186460
|
||||
141.495384 43.378901
|
||||
141.462528 43.557260
|
||||
141.462528 43.716845
|
||||
141.596297 43.810718
|
||||
141.727720 44.043055
|
||||
141.739454 44.303554
|
||||
141.805166 44.547624
|
||||
141.870877 44.721290
|
||||
141.805166 44.862100
|
||||
141.727720 45.035766
|
||||
141.683130 45.160148
|
||||
141.694865 45.284531
|
||||
141.793432 45.354936
|
||||
141.903733 45.432381
|
||||
141.992913 45.462890
|
||||
142.124335 45.376057
|
||||
142.136070 45.361976
|
||||
142.180659 45.284531
|
||||
142.347285 45.190657
|
||||
142.490442 45.042807
|
||||
142.610130 44.918424
|
||||
142.765021 44.768227
|
||||
142.931647 44.643845
|
||||
143.140515 44.493647
|
||||
143.339996 44.383346
|
||||
143.516009 44.280085
|
||||
143.736611 44.193252
|
||||
143.936092 44.129888
|
||||
144.144960 44.082951
|
||||
144.332707 44.059483
|
||||
144.443009 43.970303
|
||||
144.654224 43.906938
|
||||
144.863092 43.939794
|
||||
145.083695 44.066523
|
||||
nan nan
|
||||
141.241925 45.230553
|
||||
141.131624 45.167189
|
||||
141.241925 45.096784
|
||||
141.340492 45.089743
|
||||
141.363961 45.230553
|
||||
141.241925 45.230553
|
||||
nan nan
|
||||
141.209070 41.372356
|
||||
141.230191 41.379397
|
||||
141.363961 41.372356
|
||||
141.485996 41.379397
|
||||
141.528239 41.280830
|
||||
141.518852 41.137673
|
||||
141.495384 40.947579
|
||||
141.518852 40.722283
|
||||
141.605685 40.501681
|
||||
141.793432 40.325668
|
||||
141.903733 40.130881
|
||||
141.960057 40.022926
|
||||
nan nan
|
||||
139.906576 39.980683
|
||||
140.082589 40.182511
|
||||
140.070855 40.435969
|
||||
140.061467 40.611982
|
||||
140.148300 40.729324
|
||||
140.282070 40.787995
|
||||
140.314926 40.787995
|
||||
140.336047 40.863093
|
||||
140.347781 41.022678
|
||||
140.413493 41.130632
|
||||
140.547262 41.179916
|
||||
140.657563 41.071962
|
||||
140.744396 40.879521
|
||||
140.821842 40.797382
|
||||
140.943877 40.931151
|
||||
141.075300 40.888908
|
||||
141.197336 40.895949
|
||||
141.319371 41.088389
|
||||
141.331105 41.245627
|
||||
141.185601 41.156448
|
||||
140.943877 41.104817
|
||||
140.878166 41.264402
|
||||
140.943877 41.428681
|
||||
141.042444 41.454496
|
||||
141.209070 41.372356
|
||||
nan nan
|
||||
132.729954 44.838632
|
||||
132.786278 44.909037
|
||||
132.828521 45.113212
|
||||
132.741689 45.244634
|
||||
132.521086 45.284531
|
||||
132.178448 45.237594
|
||||
132.089268 45.120252
|
||||
132.199570 44.941893
|
||||
132.199570 44.777614
|
||||
132.300484 44.667313
|
||||
132.476496 44.596908
|
||||
132.697099 44.681394
|
||||
132.729954 44.838632
|
||||
nan nan
|
||||
139.676457 48.000000
|
||||
139.620262 47.962270
|
||||
139.420781 47.814419
|
||||
139.265890 47.619632
|
||||
139.122733 47.441272
|
||||
138.934986 47.291075
|
||||
138.747240 47.140877
|
||||
138.592349 46.960171
|
||||
138.538371 46.817014
|
||||
138.460926 46.643348
|
||||
138.404602 46.521313
|
||||
138.261445 46.324179
|
||||
138.139409 46.162247
|
||||
138.261445 46.392237
|
||||
138.172265 46.256120
|
||||
138.085432 46.110617
|
||||
137.885951 45.979194
|
||||
137.787384 45.817262
|
||||
137.587903 45.655331
|
||||
137.379035 45.509827
|
||||
137.212409 45.361976
|
||||
137.012928 45.230553
|
||||
136.869772 45.113212
|
||||
136.804060 45.035766
|
||||
136.682025 44.901997
|
||||
136.538868 44.808123
|
||||
136.395711 44.643845
|
||||
136.297144 44.500688
|
||||
136.097663 44.413855
|
||||
135.921650 44.289473
|
||||
135.776146 44.146316
|
||||
135.755025 44.033668
|
||||
135.632989 43.899898
|
||||
135.513301 43.763781
|
||||
135.346676 43.597156
|
||||
135.093217 43.404716
|
||||
134.872615 43.282680
|
||||
134.628544 43.162992
|
||||
134.330496 43.008101
|
||||
134.065304 42.846169
|
||||
133.877557 42.782804
|
||||
133.746134 42.782804
|
||||
133.546653 42.740561
|
||||
133.326051 42.677197
|
||||
133.105448 42.700665
|
||||
132.995147 42.749949
|
||||
132.929435 42.740561
|
||||
132.762810 42.829741
|
||||
132.563329 42.855556
|
||||
132.410785 42.846169
|
||||
132.377929 42.895452
|
||||
132.366195 43.033916
|
||||
132.399051 43.137177
|
||||
132.399051 43.242784
|
||||
132.288749 43.209928
|
||||
132.068147 43.090240
|
||||
132.002435 43.146564
|
||||
132.089268 43.282680
|
||||
131.934377 43.299108
|
||||
131.913256 43.418797
|
||||
131.847544 43.292068
|
||||
131.781833 43.106668
|
||||
131.648063 42.984632
|
||||
131.558884 42.928308
|
||||
131.504906 42.822701
|
||||
131.382871 42.740561
|
||||
131.284304 42.611485
|
||||
131.073089 42.611485
|
||||
130.908810 42.620873
|
||||
130.753919 42.595058
|
||||
130.810243 42.529346
|
||||
130.843099 42.423739
|
||||
130.622496 42.562202
|
||||
nan nan
|
||||
140.028612 42.120997
|
||||
139.883108 42.212523
|
||||
139.894842 42.334559
|
||||
139.927698 42.522306
|
||||
nan nan
|
||||
130.699942 42.374455
|
||||
130.732798 42.278235
|
||||
130.610762 42.301703
|
||||
130.476993 42.268848
|
||||
130.324448 42.130384
|
||||
130.103846 41.956718
|
||||
nan nan
|
||||
130.103846 41.956718
|
||||
130.000000 41.866852
|
File diff suppressed because it is too large
Load Diff
File diff suppressed because it is too large
Load Diff
@ -1,16 +0,0 @@
|
||||
%SEA Surface elevation dataset used in WAT version 1.1.
|
||||
%
|
||||
% CALL: xn = load('sea.dat');
|
||||
%
|
||||
% Size : 9524 X 2
|
||||
% Sampling Rate : 4.0 Hz
|
||||
% Device : unknown
|
||||
% Source : unknown
|
||||
% Format : ascii, c1: time c2: surface elevation
|
||||
% Description :
|
||||
% The wave data was used in one of WAFO predecessors, i.e. the Wave
|
||||
% Analysis Toolbox version 1.1 (WAT)
|
||||
%
|
||||
% Hm0 = 1.9m, Tm02 = 4.0s, Tp2 = 11.5s Tp1=5.6s
|
||||
%
|
||||
% See also
|
@ -1,144 +0,0 @@
|
||||
0.0000000e+00 1.2400000e+01 2.0700000e+02
|
||||
1.6666667e-01 1.1800000e+01 2.0480000e+02
|
||||
3.3333333e-01 1.1300000e+01 1.9940000e+02
|
||||
5.0000000e-01 1.1800000e+01 1.9560000e+02
|
||||
6.6666667e-01 1.3000000e+01 1.9500000e+02
|
||||
8.3333333e-01 1.3400000e+01 1.9550000e+02
|
||||
1.0000000e+00 1.2000000e+01 1.9250000e+02
|
||||
1.1666667e+00 1.3300000e+01 1.9250000e+02
|
||||
1.3333333e+00 1.4900000e+01 1.9450000e+02
|
||||
1.5000000e+00 1.4400000e+01 1.9790000e+02
|
||||
1.6666667e+00 1.3600000e+01 1.9800000e+02
|
||||
1.8333333e+00 1.2400000e+01 1.8510000e+02
|
||||
2.0000000e+00 1.5100000e+01 2.0260000e+02
|
||||
2.1666667e+00 1.5000000e+01 2.0240000e+02
|
||||
2.3333333e+00 1.4400000e+01 2.0540000e+02
|
||||
2.5000000e+00 1.4600000e+01 2.0430000e+02
|
||||
2.6666667e+00 1.2800000e+01 2.0370000e+02
|
||||
2.8333333e+00 1.2600000e+01 2.0620000e+02
|
||||
3.0000000e+00 1.5500000e+01 2.1350000e+02
|
||||
3.1666667e+00 1.6700000e+01 2.1370000e+02
|
||||
3.3333333e+00 1.7100000e+01 2.1490000e+02
|
||||
3.5000000e+00 1.6500000e+01 2.1430000e+02
|
||||
3.6666667e+00 1.6800000e+01 2.1440000e+02
|
||||
3.8333333e+00 1.5100000e+01 2.1060000e+02
|
||||
4.0000000e+00 1.5600000e+01 2.1050000e+02
|
||||
4.1666667e+00 1.4700000e+01 2.0960000e+02
|
||||
4.3333333e+00 1.3900000e+01 2.0860000e+02
|
||||
4.5000000e+00 1.4200000e+01 2.0660000e+02
|
||||
4.6666667e+00 1.4100000e+01 2.0500000e+02
|
||||
4.8333333e+00 1.4800000e+01 2.0500000e+02
|
||||
5.0000000e+00 1.4600000e+01 2.0050000e+02
|
||||
5.1666667e+00 1.5100000e+01 1.9520000e+02
|
||||
5.3333333e+00 1.5000000e+01 1.9360000e+02
|
||||
5.5000000e+00 1.5600000e+01 1.9170000e+02
|
||||
5.6666667e+00 1.6900000e+01 1.8940000e+02
|
||||
5.8333333e+00 1.5800000e+01 1.8690000e+02
|
||||
6.0000000e+00 1.6500000e+01 1.8230000e+02
|
||||
6.1666667e+00 1.8100000e+01 1.8330000e+02
|
||||
6.3333333e+00 1.8400000e+01 1.8080000e+02
|
||||
6.5000000e+00 1.7900000e+01 1.7960000e+02
|
||||
6.6666667e+00 1.8900000e+01 1.7910000e+02
|
||||
6.8333333e+00 2.0000000e+01 1.7830000e+02
|
||||
7.0000000e+00 2.0000000e+01 1.7830000e+02
|
||||
7.1666667e+00 2.1400000e+01 1.7780000e+02
|
||||
7.3333333e+00 2.0800000e+01 1.7530000e+02
|
||||
7.5000000e+00 2.1800000e+01 1.7380000e+02
|
||||
7.6666667e+00 2.1800000e+01 1.7130000e+02
|
||||
7.8333333e+00 2.2600000e+01 1.7220000e+02
|
||||
8.0000000e+00 2.2700000e+01 1.7260000e+02
|
||||
8.1666667e+00 2.4100000e+01 1.7300000e+02
|
||||
8.3333333e+00 2.3500000e+01 1.7180000e+02
|
||||
8.5000000e+00 2.4100000e+01 1.7080000e+02
|
||||
8.6666667e+00 2.4400000e+01 1.6950000e+02
|
||||
8.8333333e+00 2.4900000e+01 1.7010000e+02
|
||||
9.0000000e+00 2.6500000e+01 1.7290000e+02
|
||||
9.1666667e+00 2.6500000e+01 1.7210000e+02
|
||||
9.3333333e+00 2.6300000e+01 1.7200000e+02
|
||||
9.5000000e+00 2.7300000e+01 1.7250000e+02
|
||||
9.6666667e+00 2.7300000e+01 1.7520000e+02
|
||||
9.8333333e+00 2.7800000e+01 1.7220000e+02
|
||||
1.0000000e+01 2.7200000e+01 1.7460000e+02
|
||||
1.0166667e+01 2.7900000e+01 1.7790000e+02
|
||||
1.0333333e+01 2.8400000e+01 1.7830000e+02
|
||||
1.0500000e+01 2.8300000e+01 1.7590000e+02
|
||||
1.0666667e+01 2.7300000e+01 1.7460000e+02
|
||||
1.0833333e+01 2.7600000e+01 1.7310000e+02
|
||||
1.1000000e+01 2.9300000e+01 1.7520000e+02
|
||||
1.1166667e+01 2.9600000e+01 1.7580000e+02
|
||||
1.1333333e+01 2.8700000e+01 1.7430000e+02
|
||||
1.1500000e+01 2.8200000e+01 1.7300000e+02
|
||||
1.1666667e+01 2.8100000e+01 1.7000000e+02
|
||||
1.1833333e+01 2.9100000e+01 1.6820000e+02
|
||||
1.2000000e+01 2.8800000e+01 1.7040000e+02
|
||||
1.2166667e+01 2.9600000e+01 1.7150000e+02
|
||||
1.2333333e+01 2.9900000e+01 1.7160000e+02
|
||||
1.2500000e+01 2.9500000e+01 1.7210000e+02
|
||||
1.2666667e+01 3.0100000e+01 1.7560000e+02
|
||||
1.2833333e+01 3.0900000e+01 1.7520000e+02
|
||||
1.3000000e+01 3.0900000e+01 1.7550000e+02
|
||||
1.3166667e+01 3.0000000e+01 1.7590000e+02
|
||||
1.3333333e+01 3.0500000e+01 1.7490000e+02
|
||||
1.3500000e+01 3.0500000e+01 1.7340000e+02
|
||||
1.3666667e+01 3.0100000e+01 1.7190000e+02
|
||||
1.3833333e+01 3.1500000e+01 1.7180000e+02
|
||||
1.4000000e+01 3.0700000e+01 1.7190000e+02
|
||||
1.4166667e+01 2.9200000e+01 1.6820000e+02
|
||||
1.4333333e+01 3.1200000e+01 1.6770000e+02
|
||||
1.4500000e+01 3.1900000e+01 1.6970000e+02
|
||||
1.4666667e+01 3.1300000e+01 1.7140000e+02
|
||||
1.4833333e+01 3.1300000e+01 1.7500000e+02
|
||||
1.5000000e+01 3.1500000e+01 1.7470000e+02
|
||||
1.5166667e+01 3.1800000e+01 1.7430000e+02
|
||||
1.5333333e+01 3.1500000e+01 1.7460000e+02
|
||||
1.5500000e+01 3.2800000e+01 1.7610000e+02
|
||||
1.5666667e+01 3.2700000e+01 1.7710000e+02
|
||||
1.5833333e+01 3.1000000e+01 1.7970000e+02
|
||||
1.6000000e+01 3.0100000e+01 1.8010000e+02
|
||||
1.6166667e+01 3.0000000e+01 1.7880000e+02
|
||||
1.6333333e+01 2.9200000e+01 1.7930000e+02
|
||||
1.6500000e+01 2.9900000e+01 1.7890000e+02
|
||||
1.6666667e+01 3.0300000e+01 1.7940000e+02
|
||||
1.6833333e+01 3.0700000e+01 1.7930000e+02
|
||||
1.7000000e+01 3.0600000e+01 1.8020000e+02
|
||||
1.7166667e+01 3.0300000e+01 1.7970000e+02
|
||||
1.7333333e+01 3.1100000e+01 1.7920000e+02
|
||||
1.7500000e+01 2.8900000e+01 1.8200000e+02
|
||||
1.7666667e+01 3.0300000e+01 1.8100000e+02
|
||||
1.7833333e+01 2.9900000e+01 1.7940000e+02
|
||||
1.8000000e+01 3.0800000e+01 1.7920000e+02
|
||||
1.8166667e+01 2.9500000e+01 1.7950000e+02
|
||||
1.8333333e+01 3.0600000e+01 1.7900000e+02
|
||||
1.8500000e+01 3.0800000e+01 1.7880000e+02
|
||||
1.8666667e+01 3.0900000e+01 1.8070000e+02
|
||||
1.8833333e+01 3.0700000e+01 1.8040000e+02
|
||||
1.9000000e+01 3.0700000e+01 1.7990000e+02
|
||||
1.9166667e+01 2.9400000e+01 1.8020000e+02
|
||||
1.9333333e+01 2.9100000e+01 1.7990000e+02
|
||||
1.9500000e+01 3.0600000e+01 1.8040000e+02
|
||||
1.9666667e+01 3.0100000e+01 1.8110000e+02
|
||||
1.9833333e+01 3.0000000e+01 1.8130000e+02
|
||||
2.0000000e+01 2.9300000e+01 1.8300000e+02
|
||||
2.0166667e+01 3.0500000e+01 1.8390000e+02
|
||||
2.0333333e+01 3.0800000e+01 1.8470000e+02
|
||||
2.0500000e+01 2.9600000e+01 1.8530000e+02
|
||||
2.0666667e+01 2.8800000e+01 1.8600000e+02
|
||||
2.0833333e+01 2.9100000e+01 1.8640000e+02
|
||||
2.1000000e+01 2.9800000e+01 1.8570000e+02
|
||||
2.1166667e+01 2.8700000e+01 1.8750000e+02
|
||||
2.1333333e+01 2.8700000e+01 1.8640000e+02
|
||||
2.1500000e+01 2.9300000e+01 1.8610000e+02
|
||||
2.1666667e+01 2.8500000e+01 1.8800000e+02
|
||||
2.1833333e+01 2.7200000e+01 1.9140000e+02
|
||||
2.2000000e+01 2.8100000e+01 1.9130000e+02
|
||||
2.2166667e+01 2.7000000e+01 1.9330000e+02
|
||||
2.2333333e+01 2.6300000e+01 1.9900000e+02
|
||||
2.2500000e+01 2.6100000e+01 2.0270000e+02
|
||||
2.2666667e+01 2.4000000e+01 2.1240000e+02
|
||||
2.2833333e+01 2.4000000e+01 2.1840000e+02
|
||||
2.3000000e+01 2.1600000e+01 2.1300000e+02
|
||||
2.3166667e+01 1.8800000e+01 2.0440000e+02
|
||||
2.3333333e+01 1.7700000e+01 1.9810000e+02
|
||||
2.3500000e+01 1.8500000e+01 1.9840000e+02
|
||||
2.3666667e+01 1.9600000e+01 1.9660000e+02
|
||||
2.3833333e+01 1.9200000e+01 1.9410000e+02
|
@ -1,40 +0,0 @@
|
||||
1.0000000000000000e+01 1.2075320000000000e+06
|
||||
1.0000000000000000e+01 1.0013290000000000e+06
|
||||
1.0000000000000000e+01 1.1642510000000000e+06
|
||||
1.0000000000000000e+01 1.0521420000000000e+06
|
||||
1.0000000000000000e+01 1.3143320000000000e+06
|
||||
1.0000000000000000e+01 9.2781600000000000e+05
|
||||
1.0000000000000000e+01 8.5991500000000000e+05
|
||||
1.0000000000000000e+01 9.8150100000000000e+05
|
||||
1.5000000000000000e+01 1.9180900000000000e+05
|
||||
1.5000000000000000e+01 3.5513600000000000e+05
|
||||
1.5000000000000000e+01 2.5113800000000000e+05
|
||||
1.5000000000000000e+01 3.2085900000000000e+05
|
||||
1.5000000000000000e+01 4.2688000000000000e+05
|
||||
1.5000000000000000e+01 1.9037600000000000e+05
|
||||
1.5000000000000000e+01 3.3071300000000000e+05
|
||||
1.5000000000000000e+01 3.1301500000000000e+05
|
||||
2.0000000000000000e+01 6.8162000000000000e+04
|
||||
2.0000000000000000e+01 1.3884800000000000e+05
|
||||
2.0000000000000000e+01 1.0781100000000000e+05
|
||||
2.0000000000000000e+01 1.4487700000000000e+05
|
||||
2.0000000000000000e+01 1.5860000000000000e+05
|
||||
2.0000000000000000e+01 8.9558000000000000e+04
|
||||
2.0000000000000000e+01 1.7458000000000000e+05
|
||||
2.0000000000000000e+01 1.1393600000000000e+05
|
||||
2.5000000000000000e+01 5.3675000000000000e+04
|
||||
2.5000000000000000e+01 6.3114000000000000e+04
|
||||
2.5000000000000000e+01 6.6566000000000000e+04
|
||||
2.5000000000000000e+01 4.6152000000000000e+04
|
||||
2.5000000000000000e+01 5.5601000000000000e+04
|
||||
2.5000000000000000e+01 6.3908000000000000e+04
|
||||
2.5000000000000000e+01 4.6279000000000000e+04
|
||||
2.5000000000000000e+01 4.3226000000000000e+04
|
||||
3.0000000000000000e+01 1.9547000000000000e+04
|
||||
3.0000000000000000e+01 2.4162000000000000e+04
|
||||
3.0000000000000000e+01 2.7696000000000000e+04
|
||||
3.0000000000000000e+01 3.5947000000000000e+04
|
||||
3.0000000000000000e+01 4.9134000000000000e+04
|
||||
3.0000000000000000e+01 2.5054000000000000e+04
|
||||
3.0000000000000000e+01 3.0502000000000000e+04
|
||||
3.0000000000000000e+01 4.1359000000000000e+04
|
Binary file not shown.
Before Width: | Height: | Size: 23 KiB |
@ -1,211 +0,0 @@
|
||||
<?xml version="1.0" encoding="UTF-8" standalone="no"?>
|
||||
<!-- Generated by Microsoft Visio 11.0, SVG Export, v1.0 wafo.svg Page-3 -->
|
||||
<svg
|
||||
xmlns:v="http://schemas.microsoft.com/visio/2003/SVGExtensions/"
|
||||
xmlns:dc="http://purl.org/dc/elements/1.1/"
|
||||
xmlns:cc="http://web.resource.org/cc/"
|
||||
xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#"
|
||||
xmlns:svg="http://www.w3.org/2000/svg"
|
||||
xmlns="http://www.w3.org/2000/svg"
|
||||
xmlns:sodipodi="http://sodipodi.sourceforge.net/DTD/sodipodi-0.dtd"
|
||||
xmlns:inkscape="http://www.inkscape.org/namespaces/inkscape"
|
||||
width="8.26772in"
|
||||
height="11.6929in"
|
||||
viewBox="0 0 595.276 841.89"
|
||||
xml:space="preserve"
|
||||
color-interpolation-filters="sRGB"
|
||||
class="st4"
|
||||
id="svg2269"
|
||||
sodipodi:version="0.32"
|
||||
inkscape:version="0.45.1"
|
||||
sodipodi:docname="wafoLogoNewWithBorder.svg"
|
||||
inkscape:output_extension="org.inkscape.output.svg.inkscape"
|
||||
sodipodi:docbase="K:\pab\matlab\wafosvn\wafo\data"
|
||||
inkscape:export-filename="K:\pab\seamocs\wafo\wafoOrig3.png"
|
||||
inkscape:export-xdpi="90"
|
||||
inkscape:export-ydpi="90"><metadata
|
||||
id="metadata2325"><rdf:RDF><cc:Work
|
||||
rdf:about=""><dc:format>image/svg+xml</dc:format><dc:type
|
||||
rdf:resource="http://purl.org/dc/dcmitype/StillImage" /></cc:Work></rdf:RDF></metadata><defs
|
||||
id="defs2323">
|
||||
<title
|
||||
id="title2275">Page-3</title>
|
||||
<v:pageProperties
|
||||
v:shadowOffsetY="-8.50394"
|
||||
v:shadowOffsetX="8.50394"
|
||||
v:drawingUnits="24"
|
||||
v:pageScale="0.0393701"
|
||||
v:drawingScale="0.0393701" />
|
||||
|
||||
|
||||
<title
|
||||
id="title2278">Sheet.1</title>
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
<title
|
||||
id="title2306">Sheet.7</title>
|
||||
|
||||
|
||||
</defs><sodipodi:namedview
|
||||
inkscape:window-height="878"
|
||||
inkscape:window-width="1280"
|
||||
inkscape:pageshadow="2"
|
||||
inkscape:pageopacity="0.0"
|
||||
guidetolerance="10.0"
|
||||
gridtolerance="10.0"
|
||||
objecttolerance="10.0"
|
||||
borderopacity="1.0"
|
||||
bordercolor="#666666"
|
||||
pagecolor="#ffffff"
|
||||
id="base"
|
||||
inkscape:zoom="1.7424566"
|
||||
inkscape:cx="353.93418"
|
||||
inkscape:cy="530.60308"
|
||||
inkscape:window-x="0"
|
||||
inkscape:window-y="88"
|
||||
inkscape:current-layer="svg2269" />
|
||||
<v:documentProperties
|
||||
v:langID="1033"
|
||||
v:viewMarkup="false" />
|
||||
|
||||
<style
|
||||
type="text/css"
|
||||
id="style2271">
|
||||
|
||||
.st1 {fill:#99ccff;stroke:#000000;stroke-linecap:round;stroke-linejoin:round;stroke-width:0.72}
|
||||
.st2 {fill:#99ccff}
|
||||
.st3 {stroke:#000000;stroke-linecap:round;stroke-linejoin:round;stroke-width:0.72}
|
||||
.st4 {fill:none;fill-rule:evenodd;font-size:12;overflow:visible;stroke-linecap:square;stroke-miterlimit:3}
|
||||
|
||||
</style>
|
||||
|
||||
|
||||
<g
|
||||
id="g2184"
|
||||
v:mID="6"
|
||||
v:groupContext="shape"
|
||||
transform="translate(353.50185,-422.77482)"
|
||||
style="fill:#195fdd;fill-opacity:1;stroke:#000000;stroke-width:1.48000216;stroke-linecap:round;stroke-linejoin:round;stroke-miterlimit:4;stroke-dasharray:none;stroke-dashoffset:0;stroke-opacity:1">
|
||||
<title
|
||||
id="title2186">Sheet.6</title>
|
||||
<path
|
||||
d="M 61.19,841.89 C 55.291356,841.88968 49.448079,840.75226 43.98,838.54 C 39.072299,836.55264 34.545515,833.73101 30.6,830.2 C 23.504453,836.54775 14.61042,840.53172 5.15,841.6 C 3.4401274,841.79314 1.7207458,841.88996 4.5062722e-016,841.89 L 61.19,841.89 z "
|
||||
class="st1"
|
||||
id="path2188"
|
||||
style="fill:#195fdd;fill-opacity:1;stroke:#000000;stroke-width:1.48000216;stroke-linecap:round;stroke-linejoin:round;stroke-miterlimit:4;stroke-dasharray:none;stroke-dashoffset:0;stroke-opacity:1" />
|
||||
</g><path
|
||||
style="fill:#195fdd;fill-opacity:1;stroke:#000000;stroke-width:1.48000216;stroke-linecap:round;stroke-linejoin:bevel;stroke-miterlimit:4;stroke-dasharray:none;stroke-dashoffset:0;stroke-opacity:1"
|
||||
id="path2308"
|
||||
class="st2"
|
||||
d="M 215.52095,416.85104 L 215.85095,416.76104 C 220.23289,418.1841 224.80414,418.93924 229.41095,419.00104 L 199.43095,419.00104 L 23.50095,419.00104 C 63.19839,418.99975 98.6955,395.07432 112.46095,359.04104 C 112.84246,357.9133 113.24586,356.79309 113.67095,355.68104 C 121.18932,335.97437 135.26527,319.45825 153.53095,308.91104 C 167.48696,300.8565 183.31742,296.6178 199.43095,296.62104 C 220.89105,296.62261 240.78205,307.86584 251.85095,326.25104 C 252.04387,326.5693 252.23387,326.88931 252.42095,327.21104 C 254.07818,330.08102 255.49952,333.08089 256.67095,336.18104 C 248.94376,330.47985 239.6229,327.34608 230.02095,327.22104 L 229.72095,327.21104 C 207.02353,327.05002 187.61763,343.50509 184.06629,365.92354 C 180.51495,388.34199 193.88502,409.98916 215.52095,416.85104 L 215.52095,416.85104 z M 229.41095,419.00104 C 229.51428,419.00139 229.61762,419.00139 229.72095,419.00104 L 229.41095,419.00104 L 229.41095,419.00104 z M 229.72095,419.00104 L 230.02095,419.00104 L 229.72095,419.00104 L 229.72095,419.00104 z " /><path
|
||||
style="fill:#195fdd;fill-opacity:1;stroke:#000000;stroke-width:1.48000216;stroke-linecap:round;stroke-linejoin:bevel;stroke-miterlimit:4;stroke-dasharray:none;stroke-dashoffset:0;stroke-opacity:1"
|
||||
id="path2310"
|
||||
class="st3"
|
||||
d="M 215.52095,416.85104 L 215.85095,416.76104 C 220.23289,418.1841 224.80414,418.93924 229.41095,419.00104 L 199.43095,419.00104 L 23.50095,419.00104 C 63.19839,418.99975 98.6955,395.07432 112.46095,359.04104 C 112.84246,357.9133 113.24586,356.79309 113.67095,355.68104 C 121.18932,335.97437 135.26527,319.45825 153.53095,308.91104 C 167.48696,300.8565 183.31742,296.6178 199.43095,296.62104 C 220.89105,296.62261 240.78205,307.86584 251.85095,326.25104 C 252.04387,326.5693 252.23387,326.88931 252.42095,327.21104 C 254.07818,330.08102 255.49952,333.08089 256.67095,336.18104 C 248.94376,330.47985 239.6229,327.34608 230.02095,327.22104 L 229.72095,327.21104 C 207.02353,327.05002 187.61763,343.50509 184.06629,365.92354 C 180.51495,388.34199 193.88502,409.98916 215.52095,416.85104 M 229.41095,419.00104 C 229.51428,419.00139 229.61762,419.00139 229.72095,419.00104 L 229.41095,419.00104 M 229.72095,419.00104 L 230.02095,419.00104 L 229.72095,419.00104" /><g
|
||||
style="stroke:#000000;stroke-width:1.48000216;stroke-linecap:round;stroke-linejoin:bevel;stroke-miterlimit:4;stroke-dasharray:none;stroke-dashoffset:0;stroke-opacity:1"
|
||||
id="shape2-2"
|
||||
v:mID="2"
|
||||
v:groupContext="shape"
|
||||
transform="translate(203.0194,-484.1574)">
|
||||
<title
|
||||
id="title2281">Sheet.2</title>
|
||||
<path
|
||||
d="M 5.46,834.97 C 3.3155278,837.00103 1.4764469,839.33188 -1.1925672e-015,841.89 L 8.57,841.89 L 28.44,841.89 C 29.731238,836.8916 31.649812,832.07668 34.15,827.56 C 23.94872,824.92546 13.109603,827.72497 5.46,834.97 L 5.46,834.97 z "
|
||||
class="st1"
|
||||
id="path2283"
|
||||
style="fill:#99ccff;stroke:#000000;stroke-width:1.48000216;stroke-linecap:round;stroke-linejoin:bevel;stroke-miterlimit:4;stroke-dasharray:none;stroke-dashoffset:0;stroke-opacity:1" />
|
||||
</g><g
|
||||
style="stroke:#000000;stroke-width:1.48000216;stroke-linecap:round;stroke-linejoin:bevel;stroke-miterlimit:4;stroke-dasharray:none;stroke-dashoffset:0;stroke-opacity:1"
|
||||
id="shape3-4"
|
||||
v:mID="3"
|
||||
v:groupContext="shape"
|
||||
transform="translate(76.5354,-422.965)">
|
||||
<title
|
||||
id="title2286">Sheet.3</title>
|
||||
<path
|
||||
d="M 61.19,796 C 61.189242,787.5946 64.646473,779.55907 70.75,773.78 C 70.155747,772.70486 69.518568,771.65401 68.84,770.63 C 66.677258,767.36928 64.106406,764.39851 61.19,761.79 C 51.454477,770.49674 45.889798,782.93907 45.89,796 C 45.891817,806.49692 49.491912,816.67602 56.09,824.84 C 55.25174,825.13694 54.400932,825.39718 53.54,825.62 C 48.525469,826.91402 43.264531,826.91402 38.25,825.62 C 33.231033,824.32414 28.623577,821.77438 24.86,818.21 C 18.758996,812.43332 15.301988,804.40192 15.3,796 C 15.298262,790.78823 16.184266,785.61424 17.92,780.7 C 18.763907,778.31145 19.804127,775.99688 21.03,773.78 C 22.693324,770.7727 24.689169,767.96175 26.98,765.4 C 28.118655,764.13034 29.327196,762.92514 30.6,761.79 C 28.24732,759.68515 25.683747,757.82865 22.95,756.25 C 21.104007,755.18451 19.186031,754.24891 17.21,753.45 C 13.469155,757.31762 10.254769,761.66175 7.65,766.37 C 5.1498118,770.88668 3.2312383,775.7016 1.94,780.7 C 0.65066517,785.69777 -0.0011772613,790.83859 -2.3665698e-015,796 C 0.0026435879,809.05744 5.5670747,821.49558 15.3,830.2 C 17.65268,832.30485 20.216253,834.16135 22.95,835.74 C 27.681348,838.47278 32.862669,840.33873 38.25,841.25 C 40.774359,841.67605 43.329941,841.89013 45.89,841.89 C 53.94722,841.8901 61.862477,839.76901 68.84,835.74 C 73.571348,838.47278 78.752669,840.33873 84.14,841.25 C 86.667652,841.6766 89.226601,841.89068 91.79,841.89 C 97.688644,841.88968 103.53192,840.75226 109,838.54 C 110.97603,837.74109 112.89401,836.80549 114.74,835.74 C 117.47027,834.16077 120.03047,832.30429 122.38,830.2 C 121.11062,829.06458 119.90542,827.85938 118.77,826.59 C 116.47917,824.02825 114.48332,821.2173 112.82,818.21 C 111.59413,815.99312 110.55391,813.67855 109.71,811.29 C 107.97538,806.37892 107.0894,801.20841 107.09,796 C 107.08826,790.78823 107.97427,785.61424 109.71,780.7 C 110.55391,778.31145 111.59413,775.99688 112.82,773.78 C 114.48332,770.7727 116.47917,767.96175 118.77,765.4 C 119.90542,764.13062 121.11062,762.92542 122.38,761.79 C 118.43448,758.25899 113.9077,755.43736 109,753.45 C 105.25915,757.31762 102.04477,761.66175 99.44,766.37 C 96.939812,770.88668 95.021238,775.7016 93.73,780.7 C 92.440665,785.69777 91.788823,790.83859 91.79,796 C 91.789666,801.15806 92.441502,806.29547 93.73,811.29 C 94.430192,814.00083 95.315606,816.66042 96.38,819.25 C 97.276438,821.43051 98.298142,823.55739 99.44,825.62 C 94.422301,826.91574 89.157699,826.91574 84.14,825.62 C 83.279068,825.39718 82.42826,825.13694 81.59,824.84 C 77.552511,823.41294 73.859549,821.15424 70.75,818.21 C 70.083968,817.57897 69.446657,816.9183 68.84,816.23 C 67.4955,814.70535 66.30629,813.05051 65.29,811.29 C 62.605686,806.64125 61.191691,801.36809 61.19,796 z "
|
||||
class="st1"
|
||||
id="path2288"
|
||||
style="fill:#99ccff;stroke:#000000;stroke-width:1.48000216;stroke-linecap:round;stroke-linejoin:bevel;stroke-miterlimit:4;stroke-dasharray:none;stroke-dashoffset:0;stroke-opacity:1" />
|
||||
</g><g
|
||||
style="stroke:#000000;stroke-width:1.48000216;stroke-linecap:round;stroke-linejoin:bevel;stroke-miterlimit:4;stroke-dasharray:none;stroke-dashoffset:0;stroke-opacity:1"
|
||||
id="shape4-6"
|
||||
v:mID="4"
|
||||
v:groupContext="shape"
|
||||
transform="translate(198.9204,-422.965)">
|
||||
<title
|
||||
id="title2291">Sheet.4</title>
|
||||
<path
|
||||
d="M 47.81,838.54 C 44.069155,834.67238 40.854769,830.32825 38.25,825.62 C 33.233333,816.5536 30.601077,806.36178 30.6,796 L 0,796 C 0.0019883187,804.40192 3.4589958,812.43332 9.56,818.21 C 7.0683158,822.71453 3.8365461,826.76777 -3.5667036e-014,830.2 C 3.9485244,833.73199 8.4787648,836.55366 13.39,838.54 C 18.858079,840.75226 24.701356,841.88968 30.6,841.89 C 36.498644,841.88968 42.341921,840.75226 47.81,838.54 L 47.81,838.54 z "
|
||||
class="st1"
|
||||
id="path2293"
|
||||
style="fill:#99ccff;stroke:#000000;stroke-width:1.48000216;stroke-linecap:round;stroke-linejoin:bevel;stroke-miterlimit:4;stroke-dasharray:none;stroke-dashoffset:0;stroke-opacity:1" />
|
||||
</g><g
|
||||
id="shape6-10"
|
||||
v:mID="6"
|
||||
v:groupContext="shape"
|
||||
transform="translate(291.16752,-422.77482)"
|
||||
style="fill:#195fdd;fill-opacity:1;stroke:#000000;stroke-width:1.48000216;stroke-linecap:round;stroke-linejoin:bevel;stroke-miterlimit:4;stroke-dasharray:none;stroke-dashoffset:0;stroke-opacity:1">
|
||||
<title
|
||||
id="title2301">Sheet.6</title>
|
||||
<path
|
||||
d="M 61.19,841.89 C 55.291356,841.88968 49.448079,840.75226 43.98,838.54 C 39.072299,836.55264 34.545515,833.73101 30.6,830.2 C 23.504453,836.54775 14.61042,840.53172 5.15,841.6 C 3.4401274,841.79314 1.7207458,841.88996 4.5062722e-016,841.89 L 61.19,841.89 z "
|
||||
class="st1"
|
||||
id="path2303"
|
||||
style="fill:#195fdd;fill-opacity:1;stroke:#000000;stroke-width:1.48000216;stroke-linecap:round;stroke-linejoin:bevel;stroke-miterlimit:4;stroke-dasharray:none;stroke-dashoffset:0;stroke-opacity:1" />
|
||||
</g><g
|
||||
style="stroke:#000000;stroke-width:1.48000216;stroke-linecap:round;stroke-linejoin:bevel;stroke-miterlimit:4;stroke-dasharray:none;stroke-dashoffset:0;stroke-opacity:1"
|
||||
id="shape8-15"
|
||||
v:mID="8"
|
||||
v:groupContext="shape"
|
||||
transform="translate(321.3044,-438.2631)">
|
||||
<title
|
||||
id="title2313">Sheet.8</title>
|
||||
<path
|
||||
d="M 0.66,804.97 C 0.22124193,807.04805 4.9337977e-005,809.16614 -8.4213233e-016,811.29 C -0.00173203,823.19594 6.9033391,834.02169 17.7,839.04 C 19.389124,839.82517 21.146421,840.45445 22.95,840.92 C 32.112714,843.28407 41.853704,841.27543 49.333795,835.47954 C 56.813885,829.68364 61.191238,820.75278 61.19,811.29 C 61.188145,801.82898 56.809491,792.901 49.329842,787.1074 C 41.850193,781.3138 32.111014,779.30637 22.95,781.67 C 11.698204,784.57362 3.0624758,793.60065 0.66,804.97 L 0.66,804.97 z "
|
||||
class="st1"
|
||||
id="path2315"
|
||||
style="fill:#99ccff;stroke:#000000;stroke-width:1.48000216;stroke-linecap:round;stroke-linejoin:bevel;stroke-miterlimit:4;stroke-dasharray:none;stroke-dashoffset:0;stroke-opacity:1" />
|
||||
</g><g
|
||||
style="stroke:#000000;stroke-width:1.48000216;stroke-linecap:round;stroke-linejoin:bevel;stroke-miterlimit:4;stroke-dasharray:none;stroke-dashoffset:0;stroke-opacity:1"
|
||||
id="shape9-17"
|
||||
v:mID="9"
|
||||
v:groupContext="shape"
|
||||
transform="translate(244.8144,-434.6517)">
|
||||
<title
|
||||
id="title2318">Sheet.9</title>
|
||||
<path
|
||||
d="M 27.18,783.49 C 30.37046,781.1256 33.986249,779.39749 37.83,778.4 C 47.926884,775.79129 58.657172,778.50496 66.3,785.6 C 66.369401,785.473 66.439402,785.34634 66.51,785.22 C 68.139028,782.30777 69.755833,780.06867 71.975351,777.57698 L 75.762541,773.51888 C 76.9704,772.83669 74.666372,773.66976 76.150703,773.26302 C 74.017571,771.68987 75.329915,772.78466 75.088378,772.5107 C 71.389442,769.29521 68.011435,767.3416 63.49,765.45 C 57.914597,763.11726 51.933717,761.90749 45.89,761.89 C 41.970674,761.87571 38.065628,762.363 34.27,763.34 C 29.730068,764.51346 25.396538,766.37502 21.42,768.86 C 19.249037,770.21628 17.195223,771.75162 15.28,773.45 C 14.00412,774.58169 12.792244,775.78356 11.65,777.05 C 9.3432484,779.60319 7.3306275,782.40744 5.65,785.41 C 4.4101058,787.62889 3.3564901,789.94684 2.5,792.34 C 0.88804139,797.29563 0.044924941,802.46899 3.7470027e-016,807.68 C 0.00072628219,815.45325 1.9758087,823.09895 5.74,829.9 C 8.2316842,834.40453 11.463454,838.45777 15.3,841.89 C 16.569383,840.75458 17.77458,839.54938 18.91,838.28 C 21.091288,835.84112 23.005505,833.17597 24.62,830.33 C 24.700768,830.1871 24.78077,830.04376 24.86,829.9 C 18.756473,824.12093 15.299242,816.0854 15.3,807.68 L 61.19,807.68 C 61.191474,802.46766 62.080862,797.29365 63.82,792.38 L 19.4,792.38 C 20.593932,790.15426 22.125421,788.1269 23.94,786.37 C 24.243724,786.06684 24.553793,785.77011 24.87,785.48 C 25.606291,784.77851 26.377257,784.11435 27.18,783.49 L 27.18,783.49 z "
|
||||
class="st1"
|
||||
id="path2320"
|
||||
style="fill:#99ccff;stroke:#000000;stroke-width:1.48000216;stroke-linecap:round;stroke-linejoin:bevel;stroke-miterlimit:4;stroke-dasharray:none;stroke-dashoffset:0;stroke-opacity:1"
|
||||
sodipodi:nodetypes="cccsccccsccssssccccssccccccccc" />
|
||||
</g><g
|
||||
id="g2190"
|
||||
v:mID="6"
|
||||
v:groupContext="shape"
|
||||
transform="translate(229.31375,-422.77483)"
|
||||
style="fill:#195fdd;fill-opacity:1;stroke:#000000;stroke-width:1.48000216;stroke-linecap:round;stroke-linejoin:round;stroke-miterlimit:4;stroke-dasharray:none;stroke-dashoffset:0;stroke-opacity:1">
|
||||
<title
|
||||
id="title2192">Sheet.6</title>
|
||||
<path
|
||||
d="M 61.19,841.89 C 55.291356,841.88968 49.448079,840.75226 43.98,838.54 C 39.072299,836.55264 34.545515,833.73101 30.6,830.2 C 23.504453,836.54775 14.61042,840.53172 5.15,841.6 C 3.4401274,841.79314 1.7207458,841.88996 4.5062722e-016,841.89 L 61.19,841.89 z "
|
||||
class="st1"
|
||||
id="path2194"
|
||||
style="fill:#195fdd;fill-opacity:1;stroke:#000000;stroke-width:1.48000216;stroke-linecap:round;stroke-linejoin:round;stroke-miterlimit:4;stroke-dasharray:none;stroke-dashoffset:0;stroke-opacity:1" />
|
||||
</g><g
|
||||
inkscape:groupmode="layer"
|
||||
id="layer1"
|
||||
inkscape:label="W"
|
||||
style="opacity:0.57471266" /><g
|
||||
inkscape:groupmode="layer"
|
||||
id="layer2"
|
||||
inkscape:label="WAVE"
|
||||
sodipodi:insensitive="true" /></svg>
|
Before Width: | Height: | Size: 16 KiB |
Binary file not shown.
Before Width: | Height: | Size: 13 KiB |
@ -1,243 +0,0 @@
|
||||
<?xml version="1.0" encoding="UTF-8" standalone="no"?>
|
||||
<!-- Generated by Microsoft Visio 11.0, SVG Export, v1.0 wafo.svg Page-3 -->
|
||||
<svg
|
||||
xmlns:v="http://schemas.microsoft.com/visio/2003/SVGExtensions/"
|
||||
xmlns:dc="http://purl.org/dc/elements/1.1/"
|
||||
xmlns:cc="http://web.resource.org/cc/"
|
||||
xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#"
|
||||
xmlns:svg="http://www.w3.org/2000/svg"
|
||||
xmlns="http://www.w3.org/2000/svg"
|
||||
xmlns:sodipodi="http://sodipodi.sourceforge.net/DTD/sodipodi-0.dtd"
|
||||
xmlns:inkscape="http://www.inkscape.org/namespaces/inkscape"
|
||||
width="8.26772in"
|
||||
height="11.6929in"
|
||||
viewBox="0 0 595.276 841.89"
|
||||
xml:space="preserve"
|
||||
color-interpolation-filters="sRGB"
|
||||
class="st4"
|
||||
id="svg2269"
|
||||
sodipodi:version="0.32"
|
||||
inkscape:version="0.45.1"
|
||||
sodipodi:docname="wafoLogoNewWithoutBorder.svg"
|
||||
inkscape:output_extension="org.inkscape.output.svg.inkscape"
|
||||
sodipodi:docbase="L:\wafosvn\wafo\data"
|
||||
inkscape:export-filename="K:\pab\seamocs\wafo\wafoOrig3.png"
|
||||
inkscape:export-xdpi="90"
|
||||
inkscape:export-ydpi="90"><metadata
|
||||
id="metadata2325"><rdf:RDF><cc:Work
|
||||
rdf:about=""><dc:format>image/svg+xml</dc:format><dc:type
|
||||
rdf:resource="http://purl.org/dc/dcmitype/StillImage" /></cc:Work></rdf:RDF></metadata><defs
|
||||
id="defs2323">
|
||||
<title
|
||||
id="title2275">Page-3</title>
|
||||
<v:pageProperties
|
||||
v:shadowOffsetY="-8.50394"
|
||||
v:shadowOffsetX="8.50394"
|
||||
v:drawingUnits="24"
|
||||
v:pageScale="0.0393701"
|
||||
v:drawingScale="0.0393701" />
|
||||
|
||||
|
||||
<title
|
||||
id="title2278">Sheet.1</title>
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
</defs><sodipodi:namedview
|
||||
inkscape:window-height="878"
|
||||
inkscape:window-width="1280"
|
||||
inkscape:pageshadow="2"
|
||||
inkscape:pageopacity="0.0"
|
||||
guidetolerance="10.0"
|
||||
gridtolerance="10.0"
|
||||
objecttolerance="10.0"
|
||||
borderopacity="1.0"
|
||||
bordercolor="#666666"
|
||||
pagecolor="#ffffff"
|
||||
id="base"
|
||||
inkscape:zoom="1.7424566"
|
||||
inkscape:cx="353.93418"
|
||||
inkscape:cy="530.60308"
|
||||
inkscape:window-x="22"
|
||||
inkscape:window-y="0"
|
||||
inkscape:current-layer="svg2269" />
|
||||
<v:documentProperties
|
||||
v:langID="1033"
|
||||
v:viewMarkup="false" />
|
||||
|
||||
<style
|
||||
type="text/css"
|
||||
id="style2271">
|
||||
|
||||
.st1 {fill:#99ccff;stroke:#000000;stroke-linecap:round;stroke-linejoin:round;stroke-width:0.72}
|
||||
.st2 {fill:#99ccff}
|
||||
.st3 {stroke:#000000;stroke-linecap:round;stroke-linejoin:round;stroke-width:0.72}
|
||||
.st4 {fill:none;fill-rule:evenodd;font-size:12;overflow:visible;stroke-linecap:square;stroke-miterlimit:3}
|
||||
|
||||
</style>
|
||||
|
||||
|
||||
<g
|
||||
id="shape7-12"
|
||||
v:mID="7"
|
||||
v:groupContext="shape"
|
||||
transform="matrix(-1,0,0,1,256.67095,-422.88896)"
|
||||
style="fill:#195fdd;fill-opacity:1;stroke:none;stroke-opacity:1"
|
||||
inkscape:export-filename="L:\wafosvn\wafo\data\wafoLogoNewWithoutBorder.png"
|
||||
inkscape:export-xdpi="90"
|
||||
inkscape:export-ydpi="90">
|
||||
<title
|
||||
id="title2306">Sheet.7</title>
|
||||
<path
|
||||
d="M 41.15,839.74 L 40.82,839.65 C 36.438057,841.07306 31.866812,841.8282 27.26,841.89 L 57.24,841.89 L 233.17,841.89 C 193.47256,841.88871 157.97545,817.96328 144.21,781.93 C 143.82849,780.80226 143.42509,779.68205 143,778.57 C 135.48163,758.86333 121.40568,742.34721 103.14,731.8 C 89.183986,723.74546 73.353531,719.50676 57.24,719.51 C 35.779898,719.51157 15.888896,730.7548 4.82,749.14 C 4.6270847,749.45826 4.4370768,749.77827 4.25,750.1 C 2.5927697,752.96998 1.1714283,755.96985 1.5743344e-014,759.07 C 7.7271908,753.36881 17.04805,750.23504 26.65,750.11 L 26.95,750.1 C 49.647419,749.93898 69.05332,766.39405 72.604659,788.8125 C 76.155997,811.23095 62.785926,832.87812 41.15,839.74 L 41.15,839.74 z M 27.26,841.89 C 27.156667,841.89035 27.053333,841.89035 26.95,841.89 L 27.26,841.89 L 27.26,841.89 z M 26.95,841.89 L 26.65,841.89 L 26.95,841.89 L 26.95,841.89 z "
|
||||
class="st2"
|
||||
id="path2308"
|
||||
style="fill:#195fdd;fill-opacity:1;stroke:none;stroke-opacity:1" />
|
||||
<path
|
||||
d="M 41.15,839.74 L 40.82,839.65 C 36.438057,841.07306 31.866812,841.8282 27.26,841.89 L 57.24,841.89 L 233.17,841.89 C 193.47256,841.88871 157.97545,817.96328 144.21,781.93 C 143.82849,780.80226 143.42509,779.68205 143,778.57 C 135.48163,758.86333 121.40568,742.34721 103.14,731.8 C 89.183986,723.74546 73.353531,719.50676 57.24,719.51 C 35.779898,719.51157 15.888896,730.7548 4.82,749.14 C 4.6270847,749.45826 4.4370768,749.77827 4.25,750.1 C 2.5927697,752.96998 1.1714283,755.96985 1.5743344e-014,759.07 C 7.7271908,753.36881 17.04805,750.23504 26.65,750.11 L 26.95,750.1 C 49.647419,749.93898 69.05332,766.39405 72.604659,788.8125 C 76.155997,811.23095 62.785926,832.87812 41.15,839.74 M 27.26,841.89 C 27.156667,841.89035 27.053333,841.89035 26.95,841.89 L 27.26,841.89 M 26.95,841.89 L 26.65,841.89 L 26.95,841.89"
|
||||
class="st3"
|
||||
id="path2310"
|
||||
style="fill:#195fdd;fill-opacity:1;stroke:none;stroke-width:0.72000003;stroke-linecap:round;stroke-linejoin:round;stroke-opacity:1" />
|
||||
</g><g
|
||||
style="stroke:none;stroke-opacity:1"
|
||||
id="shape2-2"
|
||||
v:mID="2"
|
||||
v:groupContext="shape"
|
||||
transform="translate(203.0194,-484.1574)"
|
||||
inkscape:export-filename="L:\wafosvn\wafo\data\wafoLogoNewWithoutBorder.png"
|
||||
inkscape:export-xdpi="90"
|
||||
inkscape:export-ydpi="90">
|
||||
<title
|
||||
id="title2281">Sheet.2</title>
|
||||
<path
|
||||
d="M 5.46,834.97 C 3.3155278,837.00103 1.4764469,839.33188 -1.1925672e-015,841.89 L 8.57,841.89 L 28.44,841.89 C 29.731238,836.8916 31.649812,832.07668 34.15,827.56 C 23.94872,824.92546 13.109603,827.72497 5.46,834.97 L 5.46,834.97 z "
|
||||
class="st1"
|
||||
id="path2283"
|
||||
style="fill:#99ccff;stroke:none;stroke-width:0.72000003;stroke-linecap:round;stroke-linejoin:round;stroke-opacity:1" />
|
||||
</g><g
|
||||
style="stroke:none;stroke-opacity:1"
|
||||
id="shape3-4"
|
||||
v:mID="3"
|
||||
v:groupContext="shape"
|
||||
transform="translate(76.5354,-422.965)"
|
||||
inkscape:export-filename="L:\wafosvn\wafo\data\wafoLogoNewWithoutBorder.png"
|
||||
inkscape:export-xdpi="90"
|
||||
inkscape:export-ydpi="90">
|
||||
<title
|
||||
id="title2286">Sheet.3</title>
|
||||
<path
|
||||
d="M 61.19,796 C 61.189242,787.5946 64.646473,779.55907 70.75,773.78 C 70.155747,772.70486 69.518568,771.65401 68.84,770.63 C 66.677258,767.36928 64.106406,764.39851 61.19,761.79 C 51.454477,770.49674 45.889798,782.93907 45.89,796 C 45.891817,806.49692 49.491912,816.67602 56.09,824.84 C 55.25174,825.13694 54.400932,825.39718 53.54,825.62 C 48.525469,826.91402 43.264531,826.91402 38.25,825.62 C 33.231033,824.32414 28.623577,821.77438 24.86,818.21 C 18.758996,812.43332 15.301988,804.40192 15.3,796 C 15.298262,790.78823 16.184266,785.61424 17.92,780.7 C 18.763907,778.31145 19.804127,775.99688 21.03,773.78 C 22.693324,770.7727 24.689169,767.96175 26.98,765.4 C 28.118655,764.13034 29.327196,762.92514 30.6,761.79 C 28.24732,759.68515 25.683747,757.82865 22.95,756.25 C 21.104007,755.18451 19.186031,754.24891 17.21,753.45 C 13.469155,757.31762 10.254769,761.66175 7.65,766.37 C 5.1498118,770.88668 3.2312383,775.7016 1.94,780.7 C 0.65066517,785.69777 -0.0011772613,790.83859 -2.3665698e-015,796 C 0.0026435879,809.05744 5.5670747,821.49558 15.3,830.2 C 17.65268,832.30485 20.216253,834.16135 22.95,835.74 C 27.681348,838.47278 32.862669,840.33873 38.25,841.25 C 40.774359,841.67605 43.329941,841.89013 45.89,841.89 C 53.94722,841.8901 61.862477,839.76901 68.84,835.74 C 73.571348,838.47278 78.752669,840.33873 84.14,841.25 C 86.667652,841.6766 89.226601,841.89068 91.79,841.89 C 97.688644,841.88968 103.53192,840.75226 109,838.54 C 110.97603,837.74109 112.89401,836.80549 114.74,835.74 C 117.47027,834.16077 120.03047,832.30429 122.38,830.2 C 121.11062,829.06458 119.90542,827.85938 118.77,826.59 C 116.47917,824.02825 114.48332,821.2173 112.82,818.21 C 111.59413,815.99312 110.55391,813.67855 109.71,811.29 C 107.97538,806.37892 107.0894,801.20841 107.09,796 C 107.08826,790.78823 107.97427,785.61424 109.71,780.7 C 110.55391,778.31145 111.59413,775.99688 112.82,773.78 C 114.48332,770.7727 116.47917,767.96175 118.77,765.4 C 119.90542,764.13062 121.11062,762.92542 122.38,761.79 C 118.43448,758.25899 113.9077,755.43736 109,753.45 C 105.25915,757.31762 102.04477,761.66175 99.44,766.37 C 96.939812,770.88668 95.021238,775.7016 93.73,780.7 C 92.440665,785.69777 91.788823,790.83859 91.79,796 C 91.789666,801.15806 92.441502,806.29547 93.73,811.29 C 94.430192,814.00083 95.315606,816.66042 96.38,819.25 C 97.276438,821.43051 98.298142,823.55739 99.44,825.62 C 94.422301,826.91574 89.157699,826.91574 84.14,825.62 C 83.279068,825.39718 82.42826,825.13694 81.59,824.84 C 77.552511,823.41294 73.859549,821.15424 70.75,818.21 C 70.083968,817.57897 69.446657,816.9183 68.84,816.23 C 67.4955,814.70535 66.30629,813.05051 65.29,811.29 C 62.605686,806.64125 61.191691,801.36809 61.19,796 z "
|
||||
class="st1"
|
||||
id="path2288"
|
||||
style="fill:#99ccff;stroke:none;stroke-width:0.72000003;stroke-linecap:round;stroke-linejoin:round;stroke-opacity:1" />
|
||||
</g><g
|
||||
style="stroke:none;stroke-opacity:1"
|
||||
id="shape4-6"
|
||||
v:mID="4"
|
||||
v:groupContext="shape"
|
||||
transform="translate(198.9204,-422.965)"
|
||||
inkscape:export-filename="L:\wafosvn\wafo\data\wafoLogoNewWithoutBorder.png"
|
||||
inkscape:export-xdpi="90"
|
||||
inkscape:export-ydpi="90">
|
||||
<title
|
||||
id="title2291">Sheet.4</title>
|
||||
<path
|
||||
d="M 47.81,838.54 C 44.069155,834.67238 40.854769,830.32825 38.25,825.62 C 33.233333,816.5536 30.601077,806.36178 30.6,796 L 0,796 C 0.0019883187,804.40192 3.4589958,812.43332 9.56,818.21 C 7.0683158,822.71453 3.8365461,826.76777 -3.5667036e-014,830.2 C 3.9485244,833.73199 8.4787648,836.55366 13.39,838.54 C 18.858079,840.75226 24.701356,841.88968 30.6,841.89 C 36.498644,841.88968 42.341921,840.75226 47.81,838.54 L 47.81,838.54 z "
|
||||
class="st1"
|
||||
id="path2293"
|
||||
style="fill:#99ccff;stroke:none;stroke-width:0.72000003;stroke-linecap:round;stroke-linejoin:round;stroke-opacity:1" />
|
||||
</g><g
|
||||
id="shape6-10"
|
||||
v:mID="6"
|
||||
v:groupContext="shape"
|
||||
transform="translate(291.16752,-422.77482)"
|
||||
style="fill:#195fdd;fill-opacity:1;stroke:none;stroke-opacity:1"
|
||||
inkscape:export-filename="L:\wafosvn\wafo\data\wafoLogoNewWithoutBorder.png"
|
||||
inkscape:export-xdpi="90"
|
||||
inkscape:export-ydpi="90">
|
||||
<title
|
||||
id="title2301">Sheet.6</title>
|
||||
<path
|
||||
d="M 61.19,841.89 C 55.291356,841.88968 49.448079,840.75226 43.98,838.54 C 39.072299,836.55264 34.545515,833.73101 30.6,830.2 C 23.504453,836.54775 14.61042,840.53172 5.15,841.6 C 3.4401274,841.79314 1.7207458,841.88996 4.5062722e-016,841.89 L 61.19,841.89 z "
|
||||
class="st1"
|
||||
id="path2303"
|
||||
style="fill:#195fdd;fill-opacity:1;stroke:none;stroke-width:0.72000003;stroke-linecap:round;stroke-linejoin:round;stroke-opacity:1" />
|
||||
</g><g
|
||||
style="stroke:none;stroke-opacity:1"
|
||||
id="shape8-15"
|
||||
v:mID="8"
|
||||
v:groupContext="shape"
|
||||
transform="translate(321.3044,-438.2631)"
|
||||
inkscape:export-filename="L:\wafosvn\wafo\data\wafoLogoNewWithoutBorder.png"
|
||||
inkscape:export-xdpi="90"
|
||||
inkscape:export-ydpi="90">
|
||||
<title
|
||||
id="title2313">Sheet.8</title>
|
||||
<path
|
||||
d="M 0.66,804.97 C 0.22124193,807.04805 4.9337977e-005,809.16614 -8.4213233e-016,811.29 C -0.00173203,823.19594 6.9033391,834.02169 17.7,839.04 C 19.389124,839.82517 21.146421,840.45445 22.95,840.92 C 32.112714,843.28407 41.853704,841.27543 49.333795,835.47954 C 56.813885,829.68364 61.191238,820.75278 61.19,811.29 C 61.188145,801.82898 56.809491,792.901 49.329842,787.1074 C 41.850193,781.3138 32.111014,779.30637 22.95,781.67 C 11.698204,784.57362 3.0624758,793.60065 0.66,804.97 L 0.66,804.97 z "
|
||||
class="st1"
|
||||
id="path2315"
|
||||
style="fill:#99ccff;stroke:none;stroke-width:0.72000003;stroke-linecap:round;stroke-linejoin:round;stroke-opacity:1" />
|
||||
</g><g
|
||||
style="stroke:none;stroke-opacity:1"
|
||||
id="shape9-17"
|
||||
v:mID="9"
|
||||
v:groupContext="shape"
|
||||
transform="translate(244.8144,-434.6517)"
|
||||
inkscape:export-filename="L:\wafosvn\wafo\data\wafoLogoNewWithoutBorder.png"
|
||||
inkscape:export-xdpi="90"
|
||||
inkscape:export-ydpi="90">
|
||||
<title
|
||||
id="title2318">Sheet.9</title>
|
||||
<path
|
||||
d="M 27.18,783.49 C 30.37046,781.1256 33.986249,779.39749 37.83,778.4 C 47.926884,775.79129 58.657172,778.50496 66.3,785.6 C 66.369401,785.473 66.439402,785.34634 66.51,785.22 C 68.139028,782.30777 69.755833,780.06867 71.975351,777.57698 L 75.762541,773.51888 C 76.9704,772.83669 74.666372,773.66976 76.150703,773.26302 C 74.017571,771.68987 75.329915,772.78466 75.088378,772.5107 C 71.389442,769.29521 68.011435,767.3416 63.49,765.45 C 57.914597,763.11726 51.933717,761.90749 45.89,761.89 C 41.970674,761.87571 38.065628,762.363 34.27,763.34 C 29.730068,764.51346 25.396538,766.37502 21.42,768.86 C 19.249037,770.21628 17.195223,771.75162 15.28,773.45 C 14.00412,774.58169 12.792244,775.78356 11.65,777.05 C 9.3432484,779.60319 7.3306275,782.40744 5.65,785.41 C 4.4101058,787.62889 3.3564901,789.94684 2.5,792.34 C 0.88804139,797.29563 0.044924941,802.46899 3.7470027e-016,807.68 C 0.00072628219,815.45325 1.9758087,823.09895 5.74,829.9 C 8.2316842,834.40453 11.463454,838.45777 15.3,841.89 C 16.569383,840.75458 17.77458,839.54938 18.91,838.28 C 21.091288,835.84112 23.005505,833.17597 24.62,830.33 C 24.700768,830.1871 24.78077,830.04376 24.86,829.9 C 18.756473,824.12093 15.299242,816.0854 15.3,807.68 L 61.19,807.68 C 61.191474,802.46766 62.080862,797.29365 63.82,792.38 L 19.4,792.38 C 20.593932,790.15426 22.125421,788.1269 23.94,786.37 C 24.243724,786.06684 24.553793,785.77011 24.87,785.48 C 25.606291,784.77851 26.377257,784.11435 27.18,783.49 L 27.18,783.49 z "
|
||||
class="st1"
|
||||
id="path2320"
|
||||
style="fill:#99ccff;stroke:none;stroke-width:0.72000003;stroke-linecap:round;stroke-linejoin:round;stroke-opacity:1"
|
||||
sodipodi:nodetypes="cccsccccsccssssccccssccccccccc" />
|
||||
</g><g
|
||||
id="g2184"
|
||||
v:mID="6"
|
||||
v:groupContext="shape"
|
||||
transform="translate(353.50185,-422.77482)"
|
||||
style="fill:#195fdd;fill-opacity:1;stroke:none;stroke-opacity:1"
|
||||
inkscape:export-filename="L:\wafosvn\wafo\data\wafoLogoNewWithoutBorder.png"
|
||||
inkscape:export-xdpi="90"
|
||||
inkscape:export-ydpi="90">
|
||||
<title
|
||||
id="title2186">Sheet.6</title>
|
||||
<path
|
||||
d="M 61.19,841.89 C 55.291356,841.88968 49.448079,840.75226 43.98,838.54 C 39.072299,836.55264 34.545515,833.73101 30.6,830.2 C 23.504453,836.54775 14.61042,840.53172 5.15,841.6 C 3.4401274,841.79314 1.7207458,841.88996 4.5062722e-016,841.89 L 61.19,841.89 z "
|
||||
class="st1"
|
||||
id="path2188"
|
||||
style="fill:#195fdd;fill-opacity:1;stroke:none;stroke-width:0.72000003;stroke-linecap:round;stroke-linejoin:round;stroke-opacity:1" />
|
||||
</g><g
|
||||
id="g2190"
|
||||
v:mID="6"
|
||||
v:groupContext="shape"
|
||||
transform="translate(229.31375,-422.77483)"
|
||||
style="fill:#195fdd;fill-opacity:1;stroke:none;stroke-opacity:1"
|
||||
inkscape:export-filename="L:\wafosvn\wafo\data\wafoLogoNewWithoutBorder.png"
|
||||
inkscape:export-xdpi="90"
|
||||
inkscape:export-ydpi="90">
|
||||
<title
|
||||
id="title2192">Sheet.6</title>
|
||||
<path
|
||||
d="M 61.19,841.89 C 55.291356,841.88968 49.448079,840.75226 43.98,838.54 C 39.072299,836.55264 34.545515,833.73101 30.6,830.2 C 23.504453,836.54775 14.61042,840.53172 5.15,841.6 C 3.4401274,841.79314 1.7207458,841.88996 4.5062722e-016,841.89 L 61.19,841.89 z "
|
||||
class="st1"
|
||||
id="path2194"
|
||||
style="fill:#195fdd;fill-opacity:1;stroke:none;stroke-width:0.72000003;stroke-linecap:round;stroke-linejoin:round;stroke-opacity:1" />
|
||||
</g><g
|
||||
inkscape:groupmode="layer"
|
||||
id="layer1"
|
||||
inkscape:label="W"
|
||||
style="opacity:0.57471266" /><g
|
||||
inkscape:groupmode="layer"
|
||||
id="layer2"
|
||||
inkscape:label="WAVE"
|
||||
sodipodi:insensitive="true" /></svg>
|
Before Width: | Height: | Size: 15 KiB |
Binary file not shown.
Before Width: | Height: | Size: 24 KiB |
File diff suppressed because it is too large
Load Diff
@ -1,310 +0,0 @@
|
||||
"""
|
||||
WAFO defintions and numenclature
|
||||
|
||||
crossings :
|
||||
cycle_pairs :
|
||||
turning_points :
|
||||
wave_amplitudes :
|
||||
wave_periods :
|
||||
waves :
|
||||
|
||||
Examples
|
||||
--------
|
||||
In order to view the documentation do the following in an ipython window:
|
||||
|
||||
import wafo.definitions as wd
|
||||
wd.crossings()
|
||||
|
||||
or
|
||||
|
||||
wd.crossings?
|
||||
|
||||
|
||||
"""
|
||||
|
||||
|
||||
def wave_amplitudes():
|
||||
r"""
|
||||
Wave amplitudes and heights definitions and nomenclature
|
||||
|
||||
Definition of wave amplitudes and wave heights
|
||||
---------------------------------------------
|
||||
|
||||
<----- Direction of wave propagation
|
||||
|
||||
|
||||
|..............c_..........|
|
||||
| /| \ |
|
||||
Hd | _/ | \ | Hu
|
||||
M | / | \ |
|
||||
/ \ | M / Ac | \_ | c_
|
||||
F \ | / \m/ | \ | / \
|
||||
------d----|---u------------------d---|---u----d------ level v
|
||||
\ | /| \ | / \L
|
||||
\_ | / | At \_|_/
|
||||
\|/..| t
|
||||
t
|
||||
|
||||
Parameters
|
||||
----------
|
||||
Ac : crest amplitude
|
||||
At : trough amplitude
|
||||
Hd : wave height as defined for down crossing waves
|
||||
Hu : wave height as defined for up crossing waves
|
||||
|
||||
See also
|
||||
--------
|
||||
waves, crossings, turning_points
|
||||
"""
|
||||
print(wave_amplitudes.__doc__)
|
||||
|
||||
|
||||
def crossings():
|
||||
r"""
|
||||
Level v crossing definitions and nomenclature
|
||||
|
||||
Definition of level v crossings
|
||||
-------------------------------
|
||||
M
|
||||
. . M M
|
||||
. . . . . .
|
||||
F d . . L
|
||||
-----------------------u-------d-------o----------------- level v
|
||||
. . . . u
|
||||
. m
|
||||
m
|
||||
|
||||
Let the letters 'm', 'M', 'F', 'L','d' and 'u' in the
|
||||
figure above denote local minimum, maximum, first value, last
|
||||
value, down- and up-crossing, respectively. The remaining
|
||||
sampled values are indicated with a '.'. Values that are identical
|
||||
with v, but do not cross the level is indicated with the letter 'o'.
|
||||
We have a level up-crossing at index, k, if
|
||||
|
||||
x(k) < v and v < x(k+1)
|
||||
or if
|
||||
x(k) == v and v < x(k+1) and x(r) < v for some di < r <= k-1
|
||||
|
||||
where di is the index to the previous downcrossing.
|
||||
Similarly there is a level down-crossing at index, k, if
|
||||
|
||||
x(k) > v and v > x(k+1)
|
||||
or if
|
||||
x(k) == v and v > x(k+1) and x(r) > v for some ui < r <= k-1
|
||||
|
||||
where ui is the index to the previous upcrossing.
|
||||
|
||||
The first (F) value is a up crossing if x(1) = v and x(2) > v.
|
||||
Similarly, it is a down crossing if x(1) = v and x(2) < v.
|
||||
|
||||
See also
|
||||
--------
|
||||
wave_periods, waves, turning_points, findcross, findtp
|
||||
"""
|
||||
print(crossings.__doc__)
|
||||
|
||||
|
||||
def cycle_pairs():
|
||||
r"""
|
||||
Cycle pairs definitions and numenclature
|
||||
|
||||
Definition of Max2min and min2Max cycle pair
|
||||
--------------------------------------------
|
||||
A min2Max cycle pair (mM) is defined as the pair of a minimum
|
||||
and the following Maximum. Similarly a Max2min cycle pair (Mm)
|
||||
is defined as the pair of a Maximum and the following minimum.
|
||||
(all turning points possibly rainflowfiltered before pairing into cycles.)
|
||||
|
||||
See also
|
||||
--------
|
||||
turning_points
|
||||
"""
|
||||
print(cycle_pairs.__doc__)
|
||||
|
||||
|
||||
def wave_periods():
|
||||
r"""
|
||||
Wave periods (lengths) definitions and nomenclature
|
||||
|
||||
Definition of wave periods (lengths)
|
||||
------------------------------------
|
||||
|
||||
|
||||
<----- Direction of wave propagation
|
||||
|
||||
<-------Tu--------->
|
||||
: :
|
||||
<---Tc-----> :
|
||||
: : : <------Tcc---->
|
||||
M : c : : : :
|
||||
/ \ : M / \_ : : c_ c
|
||||
F \ :/ \m/ \: :/ \ / \
|
||||
------d--------u----------d-------u----d--------u---d-------- level v
|
||||
\ / \ / :\_ _/: :\_ L
|
||||
\_ / \_t_/ : \t_/ : : \m/
|
||||
\t/ : : : :
|
||||
: : <---Tt---> :
|
||||
<--------Ttt-------> : :
|
||||
<-----Td----->
|
||||
Tu = Up crossing period
|
||||
Td = Down crossing period
|
||||
Tc = Crest period, i.e., period between up crossing and
|
||||
the next down crossing
|
||||
Tt = Trough period, i.e., period between down crossing and
|
||||
the next up crossing
|
||||
Ttt = Trough2trough period
|
||||
Tcc = Crest2crest period
|
||||
|
||||
|
||||
<----- Direction of wave propagation
|
||||
|
||||
<--Tcf-> Tuc
|
||||
: : <-Tcb-> <->
|
||||
M : c : : : :
|
||||
/ \ : M / \_ c_ : : c
|
||||
F \ :/ \m/ \ / \___: :/ \
|
||||
------d---------u----------d---------u-------d--------u---d------ level v
|
||||
:\_ / \ __/: \_ _/ \_ L
|
||||
: \_ / \_t_/ : \t_/ \m/
|
||||
: \t/ : :
|
||||
: : : :
|
||||
<-Ttf-> <-Ttb->
|
||||
|
||||
|
||||
Tcf = Crest front period, i.e., period between up crossing and crest
|
||||
Tcb = Crest back period, i.e., period between crest and down crossing
|
||||
Ttf = Trough front period, i.e., period between down crossing and trough
|
||||
Ttb = Trough back period, i.e., period between trough and up crossing
|
||||
Also note that Tcf and Ttf can also be abbreviated by their crossing
|
||||
marker, e.g. Tuc (u2c) and Tdt (d2t), respectively. Similar applies
|
||||
to all the other wave periods and wave lengths.
|
||||
|
||||
(The nomenclature for wave length is similar, just substitute T and
|
||||
period with L and length, respectively)
|
||||
|
||||
<----- Direction of wave propagation
|
||||
|
||||
<--TMm-->
|
||||
<-TmM-> : :
|
||||
M : : M :
|
||||
/ \ : M /:\_ : M_ M
|
||||
F \ : / \m/ : \ : /: \ / \
|
||||
\ : / : \ : / : \ / \
|
||||
\ : / : \ : / : \_ _/ \_ L
|
||||
\_ : / : \_m_/ : \m_/ \m/
|
||||
\m/ : : : :
|
||||
<-----TMM-----> <----Tmm----->
|
||||
|
||||
|
||||
TmM = Period between minimum and the following Maximum
|
||||
TMm = Period between Maximum and the following minimum
|
||||
TMM = Period between Maximum and the following Maximum
|
||||
Tmm = Period between minimum and the following minimum
|
||||
|
||||
See also
|
||||
--------
|
||||
waves,
|
||||
wave_amplitudes,
|
||||
crossings,
|
||||
turning_points
|
||||
"""
|
||||
print(wave_periods.__doc__)
|
||||
|
||||
|
||||
def turning_points():
|
||||
r"""
|
||||
Turning points definitions and numenclature
|
||||
|
||||
Definition of turningpoints
|
||||
---------------------------
|
||||
<----- Direction of wave propagation
|
||||
|
||||
M M
|
||||
/ \ .... M /:\_ M_ M
|
||||
F \ | / \m/ : \ /: \ / \
|
||||
\ h | / : \ / : \ / \
|
||||
\ | / : \ / : \_ _/ \_ L
|
||||
\_ | / : \_m_/ : \m_/ \m/
|
||||
\m/ : : : :
|
||||
<------Mw-----> <-----mw----->
|
||||
|
||||
Local minimum or maximum are indicated with the
|
||||
letters 'm' or 'M'. Turning points in this connection are all
|
||||
local max (M) and min (m) and the last (L) value and the
|
||||
first (F) value if the first local extremum is a max.
|
||||
|
||||
(This choice is made in order to get the exact up-crossing intensity
|
||||
from rfc by mm2lc(tp2mm(rfc)) )
|
||||
|
||||
|
||||
See also
|
||||
--------
|
||||
waves,
|
||||
crossings,
|
||||
cycle_pairs
|
||||
findtp
|
||||
|
||||
"""
|
||||
print(turning_points.__doc__)
|
||||
|
||||
|
||||
def waves():
|
||||
r"""
|
||||
Wave definitions and nomenclature
|
||||
|
||||
Definition of trough and crest
|
||||
------------------------------
|
||||
A trough (t) is defined as the global minimum between a
|
||||
level v down-crossing (d) and the next up-crossing (u)
|
||||
and a crest (c) is defined as the global maximum between a
|
||||
level v up-crossing and the following down-crossing.
|
||||
|
||||
Definition of down- and up -crossing waves
|
||||
------------------------------------------
|
||||
A level v-down-crossing wave (dw) is a wave from a
|
||||
down-crossing to the following down-crossing.
|
||||
Similarly, a level v-up-crossing wave (uw) is a wave from an up-crossing
|
||||
to the next up-crossing.
|
||||
|
||||
Definition of trough and crest waves
|
||||
------------------------------------
|
||||
A trough-to-trough wave (tw) is a wave from a trough (t) to the
|
||||
following trough. The crest-to-crest wave (cw) is defined similarly.
|
||||
|
||||
|
||||
Definition of min2min and Max2Max wave
|
||||
--------------------------------------
|
||||
A min2min wave (mw) is defined starting from a minimum (m) and
|
||||
ending in the following minimum.
|
||||
Similarly a Max2Max wave (Mw) is thus a wave from a maximum (M)
|
||||
to the next maximum (all waves optionally rainflow filtered).
|
||||
|
||||
<----- Direction of wave propagation
|
||||
|
||||
|
||||
<------Mw-----> <----mw---->
|
||||
M : : c :
|
||||
/ \ M : / \_ : c_ c
|
||||
F \ / \m/ \ : /: \ /:\
|
||||
------d--------u----------d-------u----d--------u---d------ level v
|
||||
\ /: \ : /: : :\_ _/ : :\_ L
|
||||
\_ / : \_t_/ : : : \t_/ : : \m/
|
||||
\t/ <-------uw---------> : <-----dw----->
|
||||
: : : :
|
||||
<--------tw--------> <------cw----->
|
||||
|
||||
(F=first value and L=last value).
|
||||
|
||||
See also
|
||||
--------
|
||||
turning_points,
|
||||
crossings,
|
||||
wave_periods
|
||||
findtc,
|
||||
findcross
|
||||
"""
|
||||
print(waves.__doc__)
|
||||
|
||||
if __name__ == '__main__':
|
||||
import doctest
|
||||
doctest.testmod(optionflags=doctest.NORMALIZE_WHITESPACE)
|
@ -1,146 +0,0 @@
|
||||
"""
|
||||
Created on 20. jan. 2011
|
||||
|
||||
@author: pab
|
||||
"""
|
||||
import numpy as np
|
||||
from numpy import exp, meshgrid
|
||||
__all__ = ['peaks', 'humps', 'magic']
|
||||
|
||||
|
||||
def _magic_odd_order(n):
|
||||
ix = np.arange(n) + 1
|
||||
J, I = np.meshgrid(ix, ix)
|
||||
A = np.mod(I + J - (n + 3) / 2, n)
|
||||
B = np.mod(I + 2 * J - 2, n)
|
||||
M = n * A + B + 1
|
||||
return M
|
||||
|
||||
|
||||
def _magic_doubly_even_order(n):
|
||||
M = np.arange(1, n * n + 1).reshape(n, n)
|
||||
ix = np.mod(np.arange(n) + 1, 4) // 2
|
||||
J, I = np.meshgrid(ix, ix)
|
||||
iz = np.flatnonzero(I == J)
|
||||
M.put(iz, n * n + 1 - M.flat[iz])
|
||||
return M
|
||||
|
||||
|
||||
def _magic_even_order(n):
|
||||
p = n // 2
|
||||
M0 = magic(p)
|
||||
M = np.hstack((np.vstack((M0, M0 + 3 * p * p)),
|
||||
np.vstack((M0 + 2 * p * p, M0 + p * p))))
|
||||
if n > 2:
|
||||
k = (n - 2) // 4
|
||||
jvec = np.hstack((np.arange(k), np.arange(n - k + 1, n)))
|
||||
for i in range(p):
|
||||
for j in jvec:
|
||||
temp = M[i][j]
|
||||
M[i][j] = M[i + p][j]
|
||||
M[i + p][j] = temp
|
||||
|
||||
i = k
|
||||
j = 0
|
||||
temp = M[i][j]
|
||||
M[i][j] = M[i + p][j]
|
||||
M[i + p][j] = temp
|
||||
j = i
|
||||
temp = M[i + p][j]
|
||||
M[i + p][j] = M[i][j]
|
||||
M[i][j] = temp
|
||||
return M
|
||||
|
||||
|
||||
def magic(n):
|
||||
"""
|
||||
Return magic square for n of any orders > 2.
|
||||
|
||||
A magic square has the property that the sum of every row and column,
|
||||
as well as both diagonals, is the same number.
|
||||
|
||||
Examples
|
||||
--------
|
||||
>>> np.allclose(magic(3),
|
||||
... [[8, 1, 6],
|
||||
... [3, 5, 7],
|
||||
... [4, 9, 2]])
|
||||
True
|
||||
|
||||
>>> np.allclose(magic(4),
|
||||
... [[16, 2, 3, 13],
|
||||
... [ 5, 11, 10, 8],
|
||||
... [ 9, 7, 6, 12],
|
||||
... [ 4, 14, 15, 1]])
|
||||
True
|
||||
|
||||
>>> np.allclose(magic(6),
|
||||
... [[35, 1, 6, 26, 19, 24],
|
||||
... [ 3, 32, 7, 21, 23, 25],
|
||||
... [31, 9, 2, 22, 27, 20],
|
||||
... [ 8, 28, 33, 17, 10, 15],
|
||||
... [30, 5, 34, 12, 14, 16],
|
||||
... [ 4, 36, 29, 13, 18, 11]])
|
||||
True
|
||||
"""
|
||||
if (n < 3):
|
||||
raise ValueError('n must be greater than 2.')
|
||||
|
||||
if np.mod(n, 2) == 1:
|
||||
return _magic_odd_order(n)
|
||||
elif np.mod(n, 4) == 0:
|
||||
return _magic_doubly_even_order(n)
|
||||
return _magic_even_order(n)
|
||||
|
||||
|
||||
def peaks(x=None, y=None, n=51):
|
||||
"""
|
||||
Return the "well" known MatLab (R) peaks function
|
||||
evaluated in the [-3,3] x,y range
|
||||
|
||||
Example
|
||||
-------
|
||||
>>> import matplotlib.pyplot as plt
|
||||
>>> x,y,z = peaks()
|
||||
|
||||
h = plt.contourf(x,y,z)
|
||||
|
||||
"""
|
||||
if x is None:
|
||||
x = np.linspace(-3, 3, n)
|
||||
if y is None:
|
||||
y = np.linspace(-3, 3, n)
|
||||
|
||||
[x1, y1] = meshgrid(x, y)
|
||||
|
||||
z = (3 * (1 - x1) ** 2 * exp(-(x1 ** 2) - (y1 + 1) ** 2) -
|
||||
10 * (x1 / 5 - x1 ** 3 - y1 ** 5) * exp(-x1 ** 2 - y1 ** 2) -
|
||||
1. / 3 * exp(-(x1 + 1) ** 2 - y1 ** 2))
|
||||
|
||||
return x1, y1, z
|
||||
|
||||
|
||||
def humps(x=None):
|
||||
"""
|
||||
Computes a function that has three roots, and some humps.
|
||||
|
||||
Example
|
||||
-------
|
||||
>>> import matplotlib.pyplot as plt
|
||||
>>> x = np.linspace(0,1)
|
||||
>>> y = humps(x)
|
||||
|
||||
h = plt.plot(x,y)
|
||||
"""
|
||||
if x is None:
|
||||
y = np.linspace(0, 1)
|
||||
else:
|
||||
y = np.asarray(x)
|
||||
|
||||
return 1.0 / ((y - 0.3) ** 2 + 0.01) + 1.0 / ((y - 0.9) ** 2 + 0.04) + \
|
||||
2 * y - 5.2
|
||||
|
||||
|
||||
if __name__ == '__main__':
|
||||
from wafo.testing import test_docstrings
|
||||
test_docstrings(__file__)
|
File diff suppressed because one or more lines are too long
File diff suppressed because one or more lines are too long
@ -1,73 +0,0 @@
|
||||
{
|
||||
"metadata": {
|
||||
"name": "WAFO Chapter 3"
|
||||
},
|
||||
"nbformat": 3,
|
||||
"nbformat_minor": 0,
|
||||
"worksheets": [
|
||||
{
|
||||
"cells": [
|
||||
{
|
||||
"cell_type": "markdown",
|
||||
"metadata": {},
|
||||
"source": [
|
||||
"CHAPTER3 Demonstrates distributions of wave characteristics\n",
|
||||
"=============================================================\n",
|
||||
"\n",
|
||||
"Chapter3 contains the commands used in Chapter3 in the tutorial.\n",
|
||||
" \n",
|
||||
"Some of the commands are edited for fast computation. \n",
|
||||
"\n",
|
||||
"Section 3.2 Estimation of wave characteristics from data\n",
|
||||
"----------------------------------------------------------\n",
|
||||
"Example 1\n",
|
||||
"~~~~~~~~~~ "
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "code",
|
||||
"collapsed": false,
|
||||
"input": [
|
||||
"speed = 'fast'\n",
|
||||
"#speed = 'slow'\n",
|
||||
"\n",
|
||||
"import wafo.data as wd\n",
|
||||
"import wafo.misc as wm\n",
|
||||
"import wafo.objects as wo\n",
|
||||
"xx = wd.sea() \n",
|
||||
"xx[:,1] = wm.detrendma(xx[:,1],len(xx))\n",
|
||||
"ts = wo.mat2timeseries(xx)\n",
|
||||
"Tcrcr, ix = ts.wave_periods(vh=0, pdef='c2c', wdef='tw', rate=8)\n",
|
||||
"Tc, ixc = ts.wave_periods(vh=0, pdef='u2d', wdef='tw', rate=8)"
|
||||
],
|
||||
"language": "python",
|
||||
"metadata": {},
|
||||
"outputs": [
|
||||
{
|
||||
"ename": "AssertionError",
|
||||
"evalue": "",
|
||||
"output_type": "pyerr",
|
||||
"traceback": [
|
||||
"\u001b[1;31m---------------------------------------------------------------------------\u001b[0m\n\u001b[1;31mAssertionError\u001b[0m Traceback (most recent call last)",
|
||||
"\u001b[1;32m<ipython-input-12-5b70e90102e6>\u001b[0m in \u001b[0;36m<module>\u001b[1;34m()\u001b[0m\n\u001b[0;32m 8\u001b[0m \u001b[0mxx\u001b[0m\u001b[1;33m[\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m,\u001b[0m\u001b[1;36m1\u001b[0m\u001b[1;33m]\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mwm\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mdetrendma\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mxx\u001b[0m\u001b[1;33m[\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m,\u001b[0m\u001b[1;36m1\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m,\u001b[0m\u001b[0mlen\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mxx\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m 9\u001b[0m \u001b[0mts\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mwo\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mmat2timeseries\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mxx\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m---> 10\u001b[1;33m \u001b[0mTcrcr\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mix\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mts\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mwave_periods\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mvh\u001b[0m\u001b[1;33m=\u001b[0m\u001b[1;36m0\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mpdef\u001b[0m\u001b[1;33m=\u001b[0m\u001b[1;34m'c2c'\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mwdef\u001b[0m\u001b[1;33m=\u001b[0m\u001b[1;34m'tw'\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mrate\u001b[0m\u001b[1;33m=\u001b[0m\u001b[1;36m8\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m 11\u001b[0m \u001b[0mTc\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mixc\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mts\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mwave_periods\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mvh\u001b[0m\u001b[1;33m=\u001b[0m\u001b[1;36m0\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mpdef\u001b[0m\u001b[1;33m=\u001b[0m\u001b[1;34m'u2d'\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mwdef\u001b[0m\u001b[1;33m=\u001b[0m\u001b[1;34m'tw'\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mrate\u001b[0m\u001b[1;33m=\u001b[0m\u001b[1;36m8\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
|
||||
"\u001b[1;32mc:\\pab\\workspace\\pywafo_svn\\pywafo\\src\\wafo\\objects.pyc\u001b[0m in \u001b[0;36mwave_periods\u001b[1;34m(self, vh, pdef, wdef, index, rate)\u001b[0m\n\u001b[0;32m 1980\u001b[0m \u001b[0mn\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mceil\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mdata\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0msize\u001b[0m \u001b[1;33m*\u001b[0m \u001b[0mrate\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m 1981\u001b[0m \u001b[0mti\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mlinspace\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0margs\u001b[0m\u001b[1;33m[\u001b[0m\u001b[1;36m0\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0margs\u001b[0m\u001b[1;33m[\u001b[0m\u001b[1;33m-\u001b[0m\u001b[1;36m1\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mn\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m-> 1982\u001b[1;33m \u001b[0mx\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mstineman_interp\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mti\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0margs\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mdata\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m 1983\u001b[0m \u001b[1;32melse\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m 1984\u001b[0m \u001b[0mx\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mdata\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
|
||||
"\u001b[1;32mC:\\Python27\\lib\\site-packages\\matplotlib\\mlab.pyc\u001b[0m in \u001b[0;36mstineman_interp\u001b[1;34m(xi, x, y, yp)\u001b[0m\n\u001b[0;32m 2932\u001b[0m \u001b[0mx\u001b[0m\u001b[1;33m=\u001b[0m\u001b[0mnp\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0masarray\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mx\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mnp\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mfloat_\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m 2933\u001b[0m \u001b[0my\u001b[0m\u001b[1;33m=\u001b[0m\u001b[0mnp\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0masarray\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0my\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mnp\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mfloat_\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m-> 2934\u001b[1;33m \u001b[1;32massert\u001b[0m \u001b[0mx\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mshape\u001b[0m \u001b[1;33m==\u001b[0m \u001b[0my\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mshape\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m 2935\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m 2936\u001b[0m \u001b[1;32mif\u001b[0m \u001b[0myp\u001b[0m \u001b[1;32mis\u001b[0m \u001b[0mNone\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
|
||||
"\u001b[1;31mAssertionError\u001b[0m: "
|
||||
]
|
||||
}
|
||||
],
|
||||
"prompt_number": 12
|
||||
},
|
||||
{
|
||||
"cell_type": "code",
|
||||
"collapsed": false,
|
||||
"input": [],
|
||||
"language": "python",
|
||||
"metadata": {},
|
||||
"outputs": []
|
||||
}
|
||||
],
|
||||
"metadata": {}
|
||||
}
|
||||
]
|
||||
}
|
File diff suppressed because one or more lines are too long
File diff suppressed because one or more lines are too long
@ -1,192 +0,0 @@
|
||||
import wafo.plotbackend.plotbackend as plt
|
||||
import numpy as np
|
||||
# pyreport -o chapter1.html chapter1.py
|
||||
|
||||
#! CHAPTER1 demonstrates some applications of WAFO
|
||||
#!================================================
|
||||
#!
|
||||
#! CHAPTER1 gives an overview through examples some of the capabilities of
|
||||
#! WAFO. WAFO is a toolbox of Matlab routines for statistical analysis and
|
||||
#! simulation of random waves and loads.
|
||||
#!
|
||||
#! The commands are edited for fast computation.
|
||||
|
||||
|
||||
#! Section 1.4 Some applications of WAFO
|
||||
#!---------------------------------------
|
||||
#! Section 1.4.1 Simulation from spectrum, estimation of spectrum
|
||||
#!~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
|
||||
#! Simulation of the sea surface from spectrum
|
||||
#! The following code generates 200 seconds of data sampled with 10Hz from
|
||||
#! the Torsethaugen spectrum
|
||||
import wafo.spectrum.models as wsm
|
||||
S = wsm.Torsethaugen(Hm0=6, Tp=8)
|
||||
S1 = S.tospecdata()
|
||||
S1.plot()
|
||||
plt.show()
|
||||
|
||||
|
||||
##
|
||||
import wafo.objects as wo
|
||||
xs = S1.sim(ns=2000, dt=0.1)
|
||||
ts = wo.mat2timeseries(xs)
|
||||
ts.plot_wave('-')
|
||||
plt.show()
|
||||
|
||||
|
||||
#! Estimation of spectrum
|
||||
#!~~~~~~~~~~~~~~~~~~~~~~~
|
||||
#! A common situation is that one wants to estimate the spectrum for wave
|
||||
#! measurements. The following code simulate 20 minutes signal sampled at 4Hz
|
||||
#! and compare the spectral estimate with the original Torsethaugen spectum.
|
||||
plt.clf()
|
||||
Fs = 4
|
||||
xs = S1.sim(ns=np.fix(20 * 60 * Fs), dt=1. / Fs)
|
||||
ts = wo.mat2timeseries(xs)
|
||||
Sest = ts.tospecdata(L=400)
|
||||
S1.plot()
|
||||
Sest.plot('--')
|
||||
plt.axis([0, 3, 0, 5])
|
||||
plt.show()
|
||||
|
||||
#! Section 1.4.2 Probability distributions of wave characteristics.
|
||||
#!~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
|
||||
#! Probability distribution of wave trough period:
|
||||
#! WAFO gives the possibility of computing the exact probability
|
||||
#! distributions for a number of characteristics given a spectral density.
|
||||
#! In the following example we study the trough period extracted from the
|
||||
#! time series and compared with the theoretical density computed with exact
|
||||
#! spectrum, S1, and the estimated spectrum, Sest.
|
||||
plt.clf()
|
||||
import wafo.misc as wm
|
||||
dtyex = S1.to_t_pdf(pdef='Tt', paramt=(0, 10, 51), nit=3)
|
||||
dtyest = Sest.to_t_pdf(pdef='Tt', paramt=(0, 10, 51), nit=3)
|
||||
|
||||
T, index = ts.wave_periods(vh=0, pdef='d2u')
|
||||
bins = wm.good_bins(T, num_bins=25, odd=True)
|
||||
wm.plot_histgrm(T, bins=bins, normed=True)
|
||||
|
||||
dtyex.plot()
|
||||
dtyest.plot('-.')
|
||||
plt.axis([0, 10, 0, 0.35])
|
||||
plt.show()
|
||||
|
||||
#! Section 1.4.3 Directional spectra
|
||||
#!~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
|
||||
#! Here are a few lines of code, which produce directional spectra
|
||||
#! with frequency independent and frequency dependent spreading.
|
||||
plt.clf()
|
||||
plotflag = 1
|
||||
Nt = 101 # number of angles
|
||||
th0 = np.pi / 2 # primary direction of waves
|
||||
Sp = 15 # spreading parameter
|
||||
|
||||
D1 = wsm.Spreading(type='cos', theta0=th0, method=None)
|
||||
D12 = wsm.Spreading(type='cos', theta0=0, method='mitsuyasu')
|
||||
|
||||
SD1 = D1.tospecdata2d(S1)
|
||||
SD12 = D12.tospecdata2d(S1)
|
||||
SD1.plot()
|
||||
SD12.plot() # linestyle='dashdot')
|
||||
plt.show()
|
||||
|
||||
#! 3D Simulation of the sea surface
|
||||
#!~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
|
||||
#! The simulations show that frequency dependent spreading leads to
|
||||
#! much more irregular surface so the orientation of waves is less
|
||||
#! transparent compared to the frequency independent case.
|
||||
#
|
||||
#! Frequency independent spreading
|
||||
#plotflag = 1; iseed = 1;
|
||||
#
|
||||
#Nx = 2 ^ 8;Ny = Nx;Nt = 1;dx = 0.5; dy = dx; dt = 0.25; fftdim = 2;
|
||||
#randn('state', iseed)
|
||||
#Y1 = seasim(SD1, Nx, Ny, Nt, dx, dy, dt, fftdim, plotflag);
|
||||
#wafostamp('', '(ER)')
|
||||
#axis('fill')
|
||||
#disp('Block = 6'), pause(pstate)
|
||||
#
|
||||
###
|
||||
## Frequency dependent spreading
|
||||
#randn('state', iseed)
|
||||
#Y12 = seasim(SD12, Nx, Ny, Nt, dx, dy, dt, fftdim, plotflag);
|
||||
#wafostamp('', '(ER)')
|
||||
#axis('fill')
|
||||
#disp('Block = 7'), pause(pstate)
|
||||
#
|
||||
#! Estimation of directional spectrum
|
||||
#!~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
|
||||
#! The figure is not shown in the Tutorial
|
||||
#
|
||||
# Nx = 3; Ny = 2; Nt = 2 ^ 12; dx = 10; dy = 10;dt = 0.5;
|
||||
# F = seasim(SD12, Nx, Ny, Nt, dx, dy, dt, 1, 0);
|
||||
# Z = permute(F.Z, [3 1 2]);
|
||||
# [X, Y] = meshgrid(F.x, F.y);
|
||||
# N = Nx * Ny;
|
||||
# types = repmat(sensortypeid('n'), N, 1);
|
||||
# bfs = ones(N, 1);
|
||||
# pos = [X(:), Y(:), zeros(N, 1)];
|
||||
# h = inf;
|
||||
# nfft = 128;
|
||||
# nt = 101;
|
||||
# SDe = dat2dspec([F.t Z(:, :)], [pos types, bfs], h, nfft, nt);
|
||||
#plotspec(SDe), hold on
|
||||
#plotspec(SD12, '--'), hold off
|
||||
#disp('Block = 8'), pause(pstate)
|
||||
|
||||
#! Section 1.4.4 Fatigue, Load cycles and Markov models.
|
||||
#! ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
|
||||
#! Switching Markow chain of turningpoints
|
||||
#! In fatigue applications the exact sample path is not important, but
|
||||
#! only the tops and bottoms of the load, called the sequence of turning
|
||||
#! points (TP). From the turning points one can extract load cycles, from
|
||||
#! which damage calculations and fatigue life predictions can be
|
||||
#! performed.
|
||||
#!
|
||||
#! The commands below computes the intensity of rainflowcycles for
|
||||
#! the Gaussian model with spectrum S1 using the Markov approximation.
|
||||
#! The rainflow cycles found in the simulated load signal are shown in the
|
||||
#! figure.
|
||||
|
||||
#clf()
|
||||
#paramu = [-6 6 61];
|
||||
#frfc = spec2cmat(S1, [], 'rfc', [], paramu);
|
||||
#pdfplot(frfc);
|
||||
#hold on
|
||||
#tp = dat2tp(xs);
|
||||
#rfc = tp2rfc(tp);
|
||||
#plot(rfc(:, 2), rfc(:, 1), '.')
|
||||
#wafostamp('', '(ER)')
|
||||
#hold off
|
||||
#disp('Block = 9'), pause(pstate)
|
||||
|
||||
#! Section 1.4.5 Extreme value statistics
|
||||
#!~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
|
||||
# Plot of yura87 data
|
||||
plt.clf()
|
||||
import wafo.data as wd
|
||||
xn = wd.yura87()
|
||||
#xn = load('yura87.dat');
|
||||
plt.subplot(211)
|
||||
plt.plot(xn[::30, 0] / 3600, xn[::30, 1], '.')
|
||||
plt.title('Water level')
|
||||
plt.ylabel('(m)')
|
||||
|
||||
#! Formation of 5 min maxima
|
||||
yura = xn[:85500, 1]
|
||||
yura = np.reshape(yura, (285, 300)).T
|
||||
maxyura = yura.max(axis=0)
|
||||
plt.subplot(212)
|
||||
plt.plot(xn[299:85500:300, 0] / 3600, maxyura, '.')
|
||||
plt.xlabel('Time (h)')
|
||||
plt.ylabel('(m)')
|
||||
plt.title('Maximum 5 min water level')
|
||||
plt.show()
|
||||
|
||||
#! Estimation of GEV for yuramax
|
||||
plt.clf()
|
||||
import wafo.stats as ws
|
||||
phat = ws.genextreme.fit2(maxyura, method='ml')
|
||||
phat.plotfitsummary()
|
||||
plt.show()
|
||||
#disp('Block = 11, Last block')
|
@ -1,327 +0,0 @@
|
||||
import wafo.plotbackend.plotbackend as plt
|
||||
import numpy as np
|
||||
|
||||
# pyreport -o chapter2.html chapter2.py
|
||||
|
||||
#! CHAPTER2 Modelling random loads and stochastic waves
|
||||
#!=======================================================
|
||||
#!
|
||||
#! Chapter2 contains the commands used in Chapter 2 of the tutorial and
|
||||
#! present some tools for analysis of random functions with
|
||||
#! respect to their correlation, spectral and distributional properties.
|
||||
#! The presentation is divided into three examples:
|
||||
#!
|
||||
#! Example1 is devoted to estimation of different parameters in the model.
|
||||
#! Example2 deals with spectral densities and
|
||||
#! Example3 presents the use of WAFO to simulate samples of a Gaussian
|
||||
#! process.
|
||||
#!
|
||||
#! Some of the commands are edited for fast computation.
|
||||
#!
|
||||
#! Section 2.1 Introduction and preliminary analysis
|
||||
#!====================================================
|
||||
#! Example 1: Sea data
|
||||
#!----------------------
|
||||
#! Observed crossings compared to the expected for Gaussian signals
|
||||
|
||||
import wafo
|
||||
import wafo.objects as wo
|
||||
xx = wafo.data.sea()
|
||||
me = xx[:, 1].mean()
|
||||
sa = xx[:, 1].std()
|
||||
xx[:, 1] -= me
|
||||
ts = wo.mat2timeseries(xx)
|
||||
tp = ts.turning_points()
|
||||
|
||||
cc = tp.cycle_pairs()
|
||||
lc = cc.level_crossings()
|
||||
lc.plot()
|
||||
plt.show()
|
||||
|
||||
#! Average number of upcrossings per time unit
|
||||
#!----------------------------------------------
|
||||
#! Next we compute the mean frequency as the average number of upcrossings
|
||||
#! per time unit of the mean level (= 0); this may require interpolation in the
|
||||
#! crossing intensity curve, as follows.
|
||||
T = xx[:, 0].max() - xx[:, 0].min()
|
||||
f0 = np.interp(0, lc.args, lc.data, 0) / T # zero up-crossing frequency
|
||||
print('f0 = %g' % f0)
|
||||
|
||||
#! Turningpoints and irregularity factor
|
||||
#!----------------------------------------
|
||||
|
||||
fm = len(tp.data) / (2 * T) # frequency of maxima
|
||||
alfa = f0 / fm # approx Tm24/Tm02
|
||||
|
||||
print('fm = %g, alpha = %g, ' % (fm, alfa))
|
||||
|
||||
#! Visually examine data
|
||||
#!------------------------
|
||||
#! We finish this section with some remarks about the quality
|
||||
#! of the measured data. Especially sea surface measurements can be
|
||||
#! of poor quality. We shall now check the quality of the dataset {\tt xx}.
|
||||
#! It is always good practice to visually examine the data
|
||||
#! before the analysis to get an impression of the quality,
|
||||
#! non-linearities and narrow-bandedness of the data.
|
||||
#! First we shall plot the data and zoom in on a specific region.
|
||||
#! A part of sea data is visualized with the following commands
|
||||
plt.clf()
|
||||
ts.plot_wave('k-', tp, '*', nfig=1, nsub=1)
|
||||
|
||||
plt.axis([0, 2, -2, 2])
|
||||
plt.show()
|
||||
|
||||
#! Finding possible spurious points
|
||||
#!------------------------------------
|
||||
#! However, if the amount of data is too large for visual examinations one
|
||||
#! could use the following criteria to find possible spurious points. One
|
||||
#! must be careful using the criteria for extremevalue analysis, because
|
||||
#! it might remove extreme waves that are OK and not spurious.
|
||||
|
||||
import wafo.misc as wm
|
||||
dt = ts.sampling_period()
|
||||
# dt = np.diff(xx[:2,0])
|
||||
dcrit = 5 * dt
|
||||
ddcrit = 9.81 / 2 * dt * dt
|
||||
zcrit = 0
|
||||
inds, indg = wm.findoutliers(ts.data, zcrit, dcrit, ddcrit, verbose=True)
|
||||
|
||||
#! Section 2.2 Frequency Modeling of Load Histories
|
||||
#!----------------------------------------------------
|
||||
#! Periodogram: Raw spectrum
|
||||
#!
|
||||
plt.clf()
|
||||
Lmax = 9500
|
||||
S = ts.tospecdata(L=Lmax)
|
||||
S.plot()
|
||||
plt.axis([0, 5, 0, 0.7])
|
||||
plt.show()
|
||||
|
||||
#! Calculate moments
|
||||
#!-------------------
|
||||
mom, text = S.moment(nr=4)
|
||||
print('sigma = %g, m0 = %g' % (sa, np.sqrt(mom[0])))
|
||||
|
||||
#! Section 2.2.1 Random functions in Spectral Domain - Gaussian processes
|
||||
#!--------------------------------------------------------------------------
|
||||
#! Smoothing of spectral estimate
|
||||
#!----------------------------------
|
||||
#! By decreasing Lmax the spectrum estimate becomes smoother.
|
||||
|
||||
plt.clf()
|
||||
Lmax0 = 200
|
||||
Lmax1 = 50
|
||||
S1 = ts.tospecdata(L=Lmax0)
|
||||
S2 = ts.tospecdata(L=Lmax1)
|
||||
S1.plot('-.')
|
||||
S2.plot()
|
||||
plt.show()
|
||||
|
||||
#! Estimated autocovariance
|
||||
#!----------------------------
|
||||
#! Obviously knowing the spectrum one can compute the covariance
|
||||
#! function. The following code will compute the covariance for the
|
||||
#! unimodal spectral density S1 and compare it with estimated
|
||||
#! covariance of the signal xx.
|
||||
plt.clf()
|
||||
Lmax = 85
|
||||
R1 = S1.tocovdata(nr=1)
|
||||
Rest = ts.tocovdata(lag=Lmax)
|
||||
R1.plot('.')
|
||||
Rest.plot()
|
||||
plt.axis([0, 25, -0.1, 0.25])
|
||||
plt.show()
|
||||
|
||||
#! We can see in Figure below that the covariance function corresponding to
|
||||
#! the spectral density S2 significantly differs from the one estimated
|
||||
#! directly from data.
|
||||
#! It can be seen in Figure above that the covariance corresponding to S1
|
||||
#! agrees much better with the estimated covariance function
|
||||
|
||||
plt.clf()
|
||||
R2 = S2.tocovdata(nr=1)
|
||||
R2.plot('.')
|
||||
Rest.plot()
|
||||
plt.show()
|
||||
|
||||
#! Section 2.2.2 Transformed Gaussian models
|
||||
#!-------------------------------------------
|
||||
#! We begin with computing skewness and kurtosis
|
||||
#! for the data set xx and compare it with the second order wave approximation
|
||||
#! proposed by Winterstein:
|
||||
import wafo.stats as ws
|
||||
rho3 = ws.skew(xx[:, 1])
|
||||
rho4 = ws.kurtosis(xx[:, 1])
|
||||
|
||||
sk, ku = S1.stats_nl(moments='sk')
|
||||
|
||||
#! Comparisons of 3 transformations
|
||||
plt.clf()
|
||||
import wafo.transform.models as wtm
|
||||
gh = wtm.TrHermite(mean=me, sigma=sa, skew=sk, kurt=ku).trdata()
|
||||
g = wtm.TrLinear(mean=me, sigma=sa).trdata() # Linear transformation
|
||||
glc, gemp = lc.trdata(mean=me, sigma=sa)
|
||||
|
||||
glc.plot('b-') # Transf. estimated from level-crossings
|
||||
gh.plot('b-.') # Hermite Transf. estimated from moments
|
||||
g.plot('r')
|
||||
plt.grid('on')
|
||||
plt.show()
|
||||
|
||||
#! Test Gaussianity of a stochastic process
|
||||
#!------------------------------------------
|
||||
#! TESTGAUSSIAN simulates e(g(u)-u) = int (g(u)-u)^2 du for Gaussian processes
|
||||
#! given the spectral density, S. The result is plotted if test0 is given.
|
||||
#! This is useful for testing if the process X(t) is Gaussian.
|
||||
#! If 95% of TEST1 is less than TEST0 then X(t) is not Gaussian at a 5% level.
|
||||
#!
|
||||
#! As we see from the figure below: none of the simulated values of test1 is
|
||||
#! above 1.00. Thus the data significantly departs from a Gaussian distribution.
|
||||
plt.clf()
|
||||
test0 = glc.dist2gauss()
|
||||
#! the following test takes time
|
||||
N = len(xx)
|
||||
test1 = S1.testgaussian(ns=N, cases=50, test0=test0)
|
||||
is_gaussian = sum(test1 > test0) > 5
|
||||
print(is_gaussian)
|
||||
plt.show()
|
||||
|
||||
#! Normalplot of data xx
|
||||
#!------------------------
|
||||
#! indicates that the underlying distribution has a "heavy" upper tail and a
|
||||
#! "light" lower tail.
|
||||
plt.clf()
|
||||
import pylab
|
||||
ws.probplot(ts.data.ravel(), dist='norm', plot=pylab)
|
||||
plt.show()
|
||||
#! Section 2.2.3 Spectral densities of sea data
|
||||
#!-----------------------------------------------
|
||||
#! Example 2: Different forms of spectra
|
||||
#!
|
||||
import wafo.spectrum.models as wsm
|
||||
plt.clf()
|
||||
Hm0 = 7
|
||||
Tp = 11
|
||||
spec = wsm.Jonswap(Hm0=Hm0, Tp=Tp).tospecdata()
|
||||
spec.plot()
|
||||
plt.show()
|
||||
|
||||
#! Directional spectrum and Encountered directional spectrum
|
||||
#! Directional spectrum
|
||||
plt.clf()
|
||||
D = wsm.Spreading('cos2s')
|
||||
Sd = D.tospecdata2d(spec)
|
||||
Sd.plot()
|
||||
plt.show()
|
||||
|
||||
|
||||
##!Encountered directional spectrum
|
||||
##!---------------------------------
|
||||
#clf()
|
||||
#Se = spec2spec(Sd,'encdir',0,10);
|
||||
#plotspec(Se), hold on
|
||||
#plotspec(Sd,1,'--'), hold off
|
||||
##!wafostamp('','(ER)')
|
||||
#disp('Block = 17'),pause(pstate)
|
||||
#
|
||||
##!#! Frequency spectra
|
||||
#clf
|
||||
#Sd1 =spec2spec(Sd,'freq');
|
||||
#Sd2 = spec2spec(Se,'enc');
|
||||
#plotspec(spec), hold on
|
||||
#plotspec(Sd1,1,'.'),
|
||||
#plotspec(Sd2),
|
||||
##!wafostamp('','(ER)')
|
||||
#hold off
|
||||
#disp('Block = 18'),pause(pstate)
|
||||
#
|
||||
##!#! Wave number spectrum
|
||||
#clf
|
||||
#Sk = spec2spec(spec,'k1d')
|
||||
#Skd = spec2spec(Sd,'k1d')
|
||||
#plotspec(Sk), hold on
|
||||
#plotspec(Skd,1,'--'), hold off
|
||||
##!wafostamp('','(ER)')
|
||||
#disp('Block = 19'),pause(pstate)
|
||||
#
|
||||
##!#! Effect of waterdepth on spectrum
|
||||
#clf
|
||||
#plotspec(spec,1,'--'), hold on
|
||||
#S20 = spec;
|
||||
#S20.S = S20.S.*phi1(S20.w,20);
|
||||
#S20.h = 20;
|
||||
#plotspec(S20), hold off
|
||||
##!wafostamp('','(ER)')
|
||||
#disp('Block = 20'),pause(pstate)
|
||||
#
|
||||
##!#! Section 2.3 Simulation of transformed Gaussian process
|
||||
##!#! Example 3: Simulation of random sea
|
||||
##! The reconstruct function replaces the spurious points of seasurface by
|
||||
##! simulated data on the basis of the remaining data and a transformed Gaussian
|
||||
##! process. As noted previously one must be careful using the criteria
|
||||
##! for finding spurious points when reconstructing a dataset, because
|
||||
##! these criteria might remove the highest and steepest waves as we can see
|
||||
##! in this plot where the spurious points is indicated with a '+' sign:
|
||||
##!
|
||||
#clf
|
||||
#[y, grec] = reconstruct(xx,inds);
|
||||
#waveplot(y,'-',xx(inds,:),'+',1,1)
|
||||
#axis([0 inf -inf inf])
|
||||
##!wafostamp('','(ER)')
|
||||
#disp('Block = 21'),pause(pstate)
|
||||
#
|
||||
##! Compare transformation (grec) from reconstructed (y)
|
||||
##! with original (glc) from (xx)
|
||||
#clf
|
||||
#trplot(g), hold on
|
||||
#plot(gemp(:,1),gemp(:,2))
|
||||
#plot(glc(:,1),glc(:,2),'-.')
|
||||
#plot(grec(:,1),grec(:,2)), hold off
|
||||
#disp('Block = 22'),pause(pstate)
|
||||
#
|
||||
##!#!
|
||||
#clf
|
||||
#L = 200;
|
||||
#x = dat2gaus(y,grec);
|
||||
#Sx = dat2spec(x,L);
|
||||
#disp('Block = 23'),pause(pstate)
|
||||
#
|
||||
##!#!
|
||||
#clf
|
||||
#dt = spec2dt(Sx)
|
||||
#Ny = fix(2*60/dt) #! = 2 minutes
|
||||
#Sx.tr = grec;
|
||||
#ysim = spec2sdat(Sx,Ny);
|
||||
#waveplot(ysim,'-')
|
||||
##!wafostamp('','(CR)')
|
||||
#disp('Block = 24'),pause(pstate)
|
||||
#
|
||||
#! Estimated spectrum compared to Torsethaugen spectrum
|
||||
#!-------------------------------------------------------
|
||||
|
||||
plt.clf()
|
||||
fp = 1.1
|
||||
dw = 0.01
|
||||
H0 = S1.characteristic('Hm0')[0]
|
||||
St = wsm.Torsethaugen(Hm0=H0,Tp=2*np.pi/fp).tospecdata(np.arange(0,5+dw/2,dw))
|
||||
S1.plot()
|
||||
St.plot('-.')
|
||||
plt.axis([0, 6, 0, 0.4])
|
||||
plt.show()
|
||||
|
||||
|
||||
#! Transformed Gaussian model compared to Gaussian model
|
||||
#!--------------------------------------------------------
|
||||
dt = St.sampling_period()
|
||||
va, sk, ku = St.stats_nl(moments='vsk')
|
||||
#sa = sqrt(va)
|
||||
gh = wtm.TrHermite(mean=me, sigma=sa, skew=sk, kurt=ku, ysigma=sa)
|
||||
|
||||
ysim_t = St.sim(ns=240, dt=0.5)
|
||||
xsim_t = ysim_t.copy()
|
||||
xsim_t[:, 1] = gh.gauss2dat(ysim_t[:, 1])
|
||||
|
||||
ts_y = wo.mat2timeseries(ysim_t)
|
||||
ts_x = wo.mat2timeseries(xsim_t)
|
||||
ts_y.plot_wave(sym1='r.', ts=ts_x, sym2='b', sigma=sa, nsub=5, nfig=1)
|
||||
plt.show()
|
@ -1,615 +0,0 @@
|
||||
from wafo.plotbackend import plotbackend as plt
|
||||
import numpy as np
|
||||
|
||||
#! CHAPTER3 Demonstrates distributions of wave characteristics
|
||||
#!=============================================================
|
||||
#!
|
||||
#! Chapter3 contains the commands used in Chapter3 in the tutorial.
|
||||
#!
|
||||
#! Some of the commands are edited for fast computation.
|
||||
#!
|
||||
#! Section 3.2 Estimation of wave characteristics from data
|
||||
#!----------------------------------------------------------
|
||||
#! Example 1
|
||||
#!~~~~~~~~~~
|
||||
|
||||
speed = 'fast'
|
||||
#speed = 'slow'
|
||||
import scipy.signal as ss
|
||||
import wafo.data as wd
|
||||
import wafo.misc as wm
|
||||
import wafo.objects as wo
|
||||
import wafo.stats as ws
|
||||
import wafo.spectrum.models as wsm
|
||||
xx = wd.sea()
|
||||
xx[:, 1] = ss.detrend(xx[:, 1])
|
||||
ts = wo.mat2timeseries(xx)
|
||||
Tcrcr, ix = ts.wave_periods(vh=0, pdef='c2c', wdef='tw', rate=8)
|
||||
Tc, ixc = ts.wave_periods(vh=0, pdef='u2d', wdef='tw', rate=8)
|
||||
|
||||
#! Histogram of crestperiod compared to the kernel density estimate
|
||||
#!~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
|
||||
import wafo.kdetools as wk
|
||||
plt.clf()
|
||||
print(Tc.mean())
|
||||
print(Tc.max())
|
||||
|
||||
t = np.linspace(0.01,8,200);
|
||||
ftc = wk.TKDE(Tc, L2=0, inc=128)
|
||||
|
||||
plt.plot(t,ftc.eval_grid(t), t, ftc.eval_grid_fast(t),'-.')
|
||||
wm.plot_histgrm(Tc, normed=True)
|
||||
plt.title('Kernel Density Estimates')
|
||||
plt.xlabel('Tc [s]')
|
||||
plt.axis([0, 8, 0, 0.5])
|
||||
plt.show()
|
||||
|
||||
#! Extreme waves - model check: the highest and steepest wave
|
||||
#!~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
|
||||
plt.clf()
|
||||
S, H = ts.wave_height_steepness(kind=0)
|
||||
indS = S.argmax()
|
||||
indH = H.argmax()
|
||||
ts.plot_sp_wave([indH, indS],'k.')
|
||||
plt.show()
|
||||
|
||||
#! Does the highest wave contradict a transformed Gaussian model?
|
||||
#!----------------------------------------------------------------
|
||||
|
||||
# TODO: Fix this
|
||||
|
||||
#clf
|
||||
#inds1 = (5965:5974)'; #! points to remove
|
||||
#Nsim = 10;
|
||||
#[y1, grec1, g2, test, tobs, mu1o, mu1oStd] = ...
|
||||
# reconstruct(xx,inds1,Nsim);
|
||||
#spwaveplot(y1,indA-10)
|
||||
#hold on
|
||||
#plot(xx(inds1,1),xx(inds1,2),'+')
|
||||
#lamb = 2.;
|
||||
#muLstd = tranproc(mu1o-lamb*mu1oStd,fliplr(grec1));
|
||||
#muUstd = tranproc(mu1o+lamb*mu1oStd,fliplr(grec1));
|
||||
#plot (y1(inds1,1), [muLstd muUstd],'b-')
|
||||
#axis([1482 1498 -1 3]),
|
||||
#wafostamp([],'(ER)')
|
||||
#disp('Block = 6'),
|
||||
#pause(pstate)
|
||||
#
|
||||
##!#! Expected value (solid) compared to data removed
|
||||
#clf
|
||||
#plot(xx(inds1,1),xx(inds1,2),'+'), hold on
|
||||
#mu = tranproc(mu1o,fliplr(grec1));
|
||||
#plot(y1(inds1,1), mu), hold off
|
||||
#disp('Block = 7'), pause(pstate)
|
||||
|
||||
#! Crest height PDF
|
||||
#!------------------
|
||||
#! Transform data so that kde works better
|
||||
plt.clf()
|
||||
wave_data = ts.wave_parameters()
|
||||
Ac = wave_data['Ac']
|
||||
L2 = 0.6
|
||||
|
||||
ws.probplot(Ac**L2, dist='norm', plot=plt)
|
||||
plt.show()
|
||||
|
||||
#!#!
|
||||
plt.clf()#
|
||||
fac = wk.TKDE(Ac,L2=L2)(np.linspace(0.01,3,200), output='plot')
|
||||
fac.plot()
|
||||
# wafostamp([],'(ER)')
|
||||
print(fac.integrate(a=0.01, b=3))
|
||||
print(fac.integrate())
|
||||
print('Block = 8'),
|
||||
# pause(pstate)
|
||||
|
||||
#!#! Empirical crest height CDF
|
||||
plt.clf()
|
||||
Fac = fac.to_cdf()
|
||||
Femp = ws.edf(Ac)
|
||||
Fac.plot()
|
||||
Femp.plot()
|
||||
plt.axis([0, 2, 0, 1])
|
||||
|
||||
#wafostamp([],'(ER)')
|
||||
#disp('Block = 9'), pause(pstate)
|
||||
|
||||
#!#! Empirical crest height CDF compared to a Transformed Rayleigh approximation
|
||||
|
||||
# facr = trraylpdf(fac.x{1},'Ac',grec1);
|
||||
# Facr = cumtrapz(facr.x{1},facr.f);
|
||||
# hold on
|
||||
# plot(facr.x{1},Facr,'.')
|
||||
# axis([1.25 2.25 0.95 1])
|
||||
# wafostamp([],'(ER)')
|
||||
# disp('Block = 10'), pause(pstate)
|
||||
|
||||
#!#! Joint pdf of crest period and crest amplitude
|
||||
plt.clf()
|
||||
Tcf = wave_data['Tcf']
|
||||
Tcb = wave_data['Tcb']
|
||||
Tc = Tcf + Tcb
|
||||
fTcAc = wk.TKDE([Tc, Ac],L2=0.5, inc=256).eval_grid_fast(output='plot')
|
||||
fTcAc.labels.labx = 'Tc [s]'
|
||||
fTcAc.labels.laby = 'Ac [m]'
|
||||
fTcAc.plot()
|
||||
plt.hold(True)
|
||||
plt.plot(Tc, Ac,'k.')
|
||||
plt.hold(False)
|
||||
plt.show()
|
||||
#wafostamp([],'(ER)')
|
||||
#disp('Block = 11'), pause(pstate)
|
||||
|
||||
#!#! Example 4: Simple wave characteristics obtained from Jonswap spectrum
|
||||
plt.clf()
|
||||
S = wsm.Jonswap(Hm0=5, Tp=10).tospecdata()
|
||||
m, mt = S.moment(nr=4, even=False)
|
||||
print(m)
|
||||
print(mt)
|
||||
# disp('Block = 12'), pause(pstate)
|
||||
|
||||
plt.clf()
|
||||
S.bandwidth(['alpha'])
|
||||
ch, Sa2, chtxt = S.characteristic(['Hm0', 'Tm02'])
|
||||
|
||||
# disp('Block = 13'), pause(pstate)
|
||||
|
||||
#!#! Section 3.3.2 Explicit form approximations of wave characteristic densities
|
||||
#!#! Longuett-Higgins model for Tc and Ac
|
||||
# plt.clf()
|
||||
# t = np.linspace(0,15,100)
|
||||
# h = np.linspace(0,6,100)
|
||||
# flh = lh83pdf(t, h, [m[0],m[1], m[2])
|
||||
# #disp('Block = 14'), pause(pstate)
|
||||
#
|
||||
# #!#! Transformed Longuett-Higgins model for Tc and Ac
|
||||
# clf
|
||||
# [sk, ku ]=spec2skew(S);
|
||||
# sa = sqrt(m(1));
|
||||
# gh = hermitetr([],[sa sk ku 0]);
|
||||
# flhg = lh83pdf(t,h,[m(1),m(2),m(3)],gh);
|
||||
# disp('Block = 15'), pause(pstate)
|
||||
|
||||
#!#! Cavanie model for Tc and Ac
|
||||
# clf
|
||||
# t = np.linspace(0,10,100);
|
||||
# h = np.linspace(0,7,100);
|
||||
# fcav = cav76pdf(t,h,[m(1) m(2) m(3) m(5)],[]);
|
||||
# disp('Block = 16'), pause(pstate)
|
||||
#
|
||||
# #!#! Example 5 Transformed Rayleigh approximation of crest- vs trough- amplitude
|
||||
# clf
|
||||
# xx = load('sea.dat');
|
||||
# x = xx;
|
||||
# x(:,2) = detrend(x(:,2));
|
||||
# SS = dat2spec2(x);
|
||||
# [sk, ku, me, si ] = spec2skew(SS);
|
||||
# gh = hermitetr([],[si sk ku me]);
|
||||
# Hs = 4*si;
|
||||
# r = (0:0.05:1.1*Hs)';
|
||||
# fac_h = trraylpdf(r,'Ac',gh);
|
||||
# fat_h = trraylpdf(r,'At',gh);
|
||||
# h = (0:0.05:1.7*Hs)';
|
||||
# facat_h = trraylpdf(h,'AcAt',gh);
|
||||
# pdfplot(fac_h)
|
||||
# hold on
|
||||
# pdfplot(fat_h,'--')
|
||||
# hold off
|
||||
# wafostamp([],'(ER)')
|
||||
# disp('Block = 17'), pause(pstate)
|
||||
#
|
||||
# #!#!
|
||||
# clf
|
||||
# TC = dat2tc(xx, me);
|
||||
# tc = tp2mm(TC);
|
||||
# Ac = tc(:,2);
|
||||
# At = -tc(:,1);
|
||||
# AcAt = Ac+At;
|
||||
# disp('Block = 18'), pause(pstate)
|
||||
#
|
||||
# #!#!
|
||||
# clf
|
||||
# Fac_h = [fac_h.x{1} cumtrapz(fac_h.x{1},fac_h.f)];
|
||||
# subplot(3,1,1)
|
||||
# Fac = plotedf(Ac,Fac_h);
|
||||
# hold on
|
||||
# plot(r,1-exp(-8*r.^2/Hs^2),'.')
|
||||
# axis([1. 2. 0.9 1])
|
||||
# title('Ac CDF')
|
||||
#
|
||||
# Fat_h = [fat_h.x{1} cumtrapz(fat_h.x{1},fat_h.f)];
|
||||
# subplot(3,1,2)
|
||||
# Fat = plotedf(At,Fat_h);
|
||||
# hold on
|
||||
# plot(r,1-exp(-8*r.^2/Hs^2),'.')
|
||||
# axis([1. 2. 0.9 1])
|
||||
# title('At CDF')
|
||||
#
|
||||
# Facat_h = [facat_h.x{1} cumtrapz(facat_h.x{1},facat_h.f)];
|
||||
# subplot(3,1,3)
|
||||
# Facat = plotedf(AcAt,Facat_h);
|
||||
# hold on
|
||||
# plot(r,1-exp(-2*r.^2/Hs^2),'.')
|
||||
# axis([1.5 3.5 0.9 1])
|
||||
# title('At+Ac CDF')
|
||||
#
|
||||
# wafostamp([],'(ER)')
|
||||
# disp('Block = 19'), pause(pstate)
|
||||
#
|
||||
# #!#! Section 3.4 Exact wave distributions in transformed Gaussian Sea
|
||||
# #!#! Section 3.4.1 Density of crest period, crest length or encountered crest period
|
||||
# clf
|
||||
# S1 = torsethaugen([],[6 8],1);
|
||||
# D1 = spreading(101,'cos',pi/2,[15],[],0);
|
||||
# D12 = spreading(101,'cos',0,[15],S1.w,1);
|
||||
# SD1 = mkdspec(S1,D1);
|
||||
# SD12 = mkdspec(S1,D12);
|
||||
# disp('Block = 20'), pause(pstate)
|
||||
#
|
||||
# #!#! Crest period
|
||||
# clf
|
||||
# tic
|
||||
# f_tc = spec2tpdf(S1,[],'Tc',[0 11 56],[],4);
|
||||
# toc
|
||||
# pdfplot(f_tc)
|
||||
# wafostamp([],'(ER)')
|
||||
# simpson(f_tc.x{1},f_tc.f)
|
||||
# disp('Block = 21'), pause(pstate)
|
||||
#
|
||||
# #!#! Crest length
|
||||
#
|
||||
# if strncmpi(speed,'slow',1)
|
||||
# opt1 = rindoptset('speed',5,'method',3);
|
||||
# opt2 = rindoptset('speed',5,'nit',2,'method',0);
|
||||
# else
|
||||
# #! fast
|
||||
# opt1 = rindoptset('speed',7,'method',3);
|
||||
# opt2 = rindoptset('speed',7,'nit',2,'method',0);
|
||||
# end
|
||||
#
|
||||
#
|
||||
# clf
|
||||
# if strncmpi(speed,'slow',1)
|
||||
# NITa = 5;
|
||||
# else
|
||||
# disp('NIT=5 may take time, running with NIT=3 in the following')
|
||||
# NITa = 3;
|
||||
# end
|
||||
# #!f_Lc = spec2tpdf2(S1,[],'Lc',[0 200 81],opt1); #! Faster and more accurate
|
||||
# f_Lc = spec2tpdf(S1,[],'Lc',[0 200 81],[],NITa);
|
||||
# pdfplot(f_Lc,'-.')
|
||||
# wafostamp([],'(ER)')
|
||||
# disp('Block = 22'), pause(pstate)
|
||||
#
|
||||
#
|
||||
# f_Lc_1 = spec2tpdf(S1,[],'Lc',[0 200 81],1.5,NITa);
|
||||
# #!f_Lc_1 = spec2tpdf2(S1,[],'Lc',[0 200 81],1.5,opt1);
|
||||
#
|
||||
# hold on
|
||||
# pdfplot(f_Lc_1)
|
||||
# wafostamp([],'(ER)')
|
||||
#
|
||||
# disp('Block = 23'), pause(pstate)
|
||||
# #!#!
|
||||
# clf
|
||||
# simpson(f_Lc.x{1},f_Lc.f)
|
||||
# simpson(f_Lc_1.x{1},f_Lc_1.f)
|
||||
#
|
||||
# disp('Block = 24'), pause(pstate)
|
||||
# #!#!
|
||||
# clf
|
||||
# tic
|
||||
#
|
||||
# f_Lc_d1 = spec2tpdf(rotspec(SD1,pi/2),[],'Lc',[0 300 121],[],NITa);
|
||||
# f_Lc_d12 = spec2tpdf(SD12,[],'Lc',[0 200 81],[],NITa);
|
||||
# #! f_Lc_d1 = spec2tpdf2(rotspec(SD1,pi/2),[],'Lc',[0 300 121],opt1);
|
||||
# #! f_Lc_d12 = spec2tpdf2(SD12,[],'Lc',[0 200 81],opt1);
|
||||
# toc
|
||||
# pdfplot(f_Lc_d1,'-.'), hold on
|
||||
# pdfplot(f_Lc_d12), hold off
|
||||
# wafostamp([],'(ER)')
|
||||
#
|
||||
# disp('Block = 25'), pause(pstate)
|
||||
#
|
||||
# #!#!
|
||||
#
|
||||
#
|
||||
# clf
|
||||
# opt1 = rindoptset('speed',5,'method',3);
|
||||
# SD1r = rotspec(SD1,pi/2);
|
||||
# if strncmpi(speed,'slow',1)
|
||||
# f_Lc_d1_5 = spec2tpdf(SD1r,[], 'Lc',[0 300 121],[],5);
|
||||
# pdfplot(f_Lc_d1_5), hold on
|
||||
# else
|
||||
# #! fast
|
||||
# disp('Run the following example only if you want a check on computing time')
|
||||
# disp('Edit the command file and remove #!')
|
||||
# end
|
||||
# f_Lc_d1_3 = spec2tpdf(SD1r,[],'Lc',[0 300 121],[],3);
|
||||
# f_Lc_d1_2 = spec2tpdf(SD1r,[],'Lc',[0 300 121],[],2);
|
||||
# f_Lc_d1_0 = spec2tpdf(SD1r,[],'Lc',[0 300 121],[],0);
|
||||
# #!f_Lc_d1_n4 = spec2tpdf2(SD1r,[],'Lc',[0 400 161],opt1);
|
||||
#
|
||||
# pdfplot(f_Lc_d1_3), hold on
|
||||
# pdfplot(f_Lc_d1_2)
|
||||
# pdfplot(f_Lc_d1_0)
|
||||
# #!pdfplot(f_Lc_d1_n4)
|
||||
#
|
||||
# #!simpson(f_Lc_d1_n4.x{1},f_Lc_d1_n4.f)
|
||||
#
|
||||
# disp('Block = 26'), pause(pstate)
|
||||
#
|
||||
# #!#! Section 3.4.2 Density of wave period, wave length or encountered wave period
|
||||
# #!#! Example 7: Crest period and high crest waves
|
||||
# clf
|
||||
# tic
|
||||
# xx = load('sea.dat');
|
||||
# x = xx;
|
||||
# x(:,2) = detrend(x(:,2));
|
||||
# SS = dat2spec(x);
|
||||
# si = sqrt(spec2mom(SS,1));
|
||||
# SS.tr = dat2tr(x);
|
||||
# Hs = 4*si
|
||||
# method = 0;
|
||||
# rate = 2;
|
||||
# [S, H, Ac, At, Tcf, Tcb, z_ind, yn] = dat2steep(x,rate,method);
|
||||
# Tc = Tcf+Tcb;
|
||||
# t = linspace(0.01,8,200);
|
||||
# ftc1 = kde(Tc,{'L2',0},t);
|
||||
# pdfplot(ftc1)
|
||||
# hold on
|
||||
# #! f_t = spec2tpdf(SS,[],'Tc',[0 8 81],0,4);
|
||||
# f_t = spec2tpdf(SS,[],'Tc',[0 8 81],0,2);
|
||||
# simpson(f_t.x{1},f_t.f)
|
||||
# pdfplot(f_t,'-.')
|
||||
# hold off
|
||||
# wafostamp([],'(ER)')
|
||||
# toc
|
||||
# disp('Block = 27'), pause(pstate)
|
||||
#
|
||||
# #!#!
|
||||
# clf
|
||||
# tic
|
||||
#
|
||||
# if strncmpi(speed,'slow',1)
|
||||
# NIT = 4;
|
||||
# else
|
||||
# NIT = 2;
|
||||
# end
|
||||
# #! f_t2 = spec2tpdf(SS,[],'Tc',[0 8 81],[Hs/2],4);
|
||||
# tic
|
||||
# f_t2 = spec2tpdf(SS,[],'Tc',[0 8 81],Hs/2,NIT);
|
||||
# toc
|
||||
#
|
||||
# Pemp = sum(Ac>Hs/2)/sum(Ac>0)
|
||||
# simpson(f_t2.x{1},f_t2.f)
|
||||
# index = find(Ac>Hs/2);
|
||||
# ftc1 = kde(Tc(index),{'L2',0},t);
|
||||
# ftc1.f = Pemp*ftc1.f;
|
||||
# pdfplot(ftc1)
|
||||
# hold on
|
||||
# pdfplot(f_t2,'-.')
|
||||
# hold off
|
||||
# wafostamp([],'(ER)')
|
||||
# toc
|
||||
# disp('Block = 28'), pause(pstate)
|
||||
#
|
||||
# #!#! Example 8: Wave period for high crest waves
|
||||
# #! clf
|
||||
# tic
|
||||
# f_tcc2 = spec2tccpdf(SS,[],'t>',[0 12 61],[Hs/2],[0],-1);
|
||||
# toc
|
||||
# simpson(f_tcc2.x{1},f_tcc2.f)
|
||||
# f_tcc3 = spec2tccpdf(SS,[],'t>',[0 12 61],[Hs/2],[0],3,5);
|
||||
# #! f_tcc3 = spec2tccpdf(SS,[],'t>',[0 12 61],[Hs/2],[0],1,5);
|
||||
# simpson(f_tcc3.x{1},f_tcc3.f)
|
||||
# pdfplot(f_tcc2,'-.')
|
||||
# hold on
|
||||
# pdfplot(f_tcc3)
|
||||
# hold off
|
||||
# toc
|
||||
# disp('Block = 29'), pause(pstate)
|
||||
#
|
||||
# #!#!
|
||||
# clf
|
||||
# [TC tc_ind v_ind] = dat2tc(yn,[],'dw');
|
||||
# N = length(tc_ind);
|
||||
# t_ind = tc_ind(1:2:N);
|
||||
# c_ind = tc_ind(2:2:N);
|
||||
# Pemp = sum(yn(t_ind,2)<-Hs/2 & yn(c_ind,2)>Hs/2)/length(t_ind)
|
||||
# ind = find(yn(t_ind,2)<-Hs/2 & yn(c_ind,2)>Hs/2);
|
||||
# spwaveplot(yn,ind(2:4))
|
||||
# wafostamp([],'(ER)')
|
||||
# disp('Block = 30'), pause(pstate)
|
||||
#
|
||||
# #!#!
|
||||
# clf
|
||||
# Tcc = yn(v_ind(1+2*ind),1)-yn(v_ind(1+2*(ind-1)),1);
|
||||
# t = linspace(0.01,14,200);
|
||||
# ftcc1 = kde(Tcc,{'kernel' 'epan','L2',0},t);
|
||||
# ftcc1.f = Pemp*ftcc1.f;
|
||||
# pdfplot(ftcc1,'-.')
|
||||
# wafostamp([],'(ER)')
|
||||
# disp('Block = 31'), pause(pstate)
|
||||
#
|
||||
# tic
|
||||
# f_tcc22_1 = spec2tccpdf(SS,[],'t>',[0 12 61],[Hs/2],[Hs/2],-1);
|
||||
# toc
|
||||
# simpson(f_tcc22_1.x{1},f_tcc22_1.f)
|
||||
# hold on
|
||||
# pdfplot(f_tcc22_1)
|
||||
# hold off
|
||||
# wafostamp([],'(ER)')
|
||||
# disp('Block = 32'), pause(pstate)
|
||||
#
|
||||
# disp('The rest of this chapter deals with joint densities.')
|
||||
# disp('Some calculations may take some time.')
|
||||
# disp('You could experiment with other NIT.')
|
||||
# #!return
|
||||
#
|
||||
# #!#! Section 3.4.3 Joint density of crest period and crest height
|
||||
# #!#! Example 9. Some preliminary analysis of the data
|
||||
# clf
|
||||
# tic
|
||||
# yy = load('gfaksr89.dat');
|
||||
# SS = dat2spec(yy);
|
||||
# si = sqrt(spec2mom(SS,1));
|
||||
# SS.tr = dat2tr(yy);
|
||||
# Hs = 4*si
|
||||
# v = gaus2dat([0 0],SS.tr);
|
||||
# v = v(2)
|
||||
# toc
|
||||
# disp('Block = 33'), pause(pstate)
|
||||
#
|
||||
# #!#!
|
||||
# clf
|
||||
# tic
|
||||
# [TC, tc_ind, v_ind] = dat2tc(yy,v,'dw');
|
||||
# N = length(tc_ind);
|
||||
# t_ind = tc_ind(1:2:N);
|
||||
# c_ind = tc_ind(2:2:N);
|
||||
# v_ind_d = v_ind(1:2:N+1);
|
||||
# v_ind_u = v_ind(2:2:N+1);
|
||||
# T_d = ecross(yy(:,1),yy(:,2),v_ind_d,v);
|
||||
# T_u = ecross(yy(:,1),yy(:,2),v_ind_u,v);
|
||||
#
|
||||
# Tc = T_d(2:end)-T_u(1:end);
|
||||
# Tt = T_u(1:end)-T_d(1:end-1);
|
||||
# Tcf = yy(c_ind,1)-T_u;
|
||||
# Ac = yy(c_ind,2)-v;
|
||||
# At = v-yy(t_ind,2);
|
||||
# toc
|
||||
# disp('Block = 34'), pause(pstate)
|
||||
#
|
||||
# #!#!
|
||||
# clf
|
||||
# tic
|
||||
# t = linspace(0.01,15,200);
|
||||
# kopt3 = kdeoptset('hs',0.25,'L2',0);
|
||||
# ftc1 = kde(Tc,kopt3,t);
|
||||
# ftt1 = kde(Tt,kopt3,t);
|
||||
# pdfplot(ftt1,'k')
|
||||
# hold on
|
||||
# pdfplot(ftc1,'k-.')
|
||||
# f_tc4 = spec2tpdf(SS,[],'Tc',[0 12 81],0,4,5);
|
||||
# f_tc2 = spec2tpdf(SS,[],'Tc',[0 12 81],0,2,5);
|
||||
# f_tc = spec2tpdf(SS,[],'Tc',[0 12 81],0,-1);
|
||||
# pdfplot(f_tc,'b')
|
||||
# hold off
|
||||
# legend('kde(Tt)','kde(Tc)','f_{tc}')
|
||||
# wafostamp([],'(ER)')
|
||||
# toc
|
||||
# disp('Block = 35'), pause(pstate)
|
||||
#
|
||||
# #!#! Example 10: Joint characteristics of a half wave:
|
||||
# #!#! position and height of a crest for a wave with given period
|
||||
# clf
|
||||
# tic
|
||||
# ind = find(4.4<Tc & Tc<4.6);
|
||||
# f_AcTcf = kde([Tcf(ind) Ac(ind)],{'L2',[1 .5]});
|
||||
# pdfplot(f_AcTcf)
|
||||
# hold on
|
||||
# plot(Tcf(ind), Ac(ind),'.');
|
||||
# wafostamp([],'(ER)')
|
||||
# toc
|
||||
# disp('Block = 36'), pause(pstate)
|
||||
#
|
||||
# #!#!
|
||||
# clf
|
||||
# tic
|
||||
# opt1 = rindoptset('speed',5,'method',3);
|
||||
# opt2 = rindoptset('speed',5,'nit',2,'method',0);
|
||||
#
|
||||
# f_tcfac1 = spec2thpdf(SS,[],'TcfAc',[4.5 4.5 46],[0:0.25:8],opt1);
|
||||
# f_tcfac2 = spec2thpdf(SS,[],'TcfAc',[4.5 4.5 46],[0:0.25:8],opt2);
|
||||
#
|
||||
# pdfplot(f_tcfac1,'-.')
|
||||
# hold on
|
||||
# pdfplot(f_tcfac2)
|
||||
# plot(Tcf(ind), Ac(ind),'.');
|
||||
#
|
||||
# simpson(f_tcfac1.x{1},simpson(f_tcfac1.x{2},f_tcfac1.f,1))
|
||||
# simpson(f_tcfac2.x{1},simpson(f_tcfac2.x{2},f_tcfac2.f,1))
|
||||
# f_tcf4=spec2tpdf(SS,[],'Tc',[4.5 4.5 46],[0:0.25:8],6);
|
||||
# f_tcf4.f(46)
|
||||
# toc
|
||||
# wafostamp([],'(ER)')
|
||||
# disp('Block = 37'), pause(pstate)
|
||||
#
|
||||
# #!#!
|
||||
# clf
|
||||
# f_tcac_s = spec2thpdf(SS,[],'TcAc',[0 12 81],[Hs/2:0.1:2*Hs],opt1);
|
||||
# disp('Block = 38'), pause(pstate)
|
||||
#
|
||||
# clf
|
||||
# tic
|
||||
# mom = spec2mom(SS,4,[],0);
|
||||
# t = f_tcac_s.x{1};
|
||||
# h = f_tcac_s.x{2};
|
||||
# flh_g = lh83pdf(t',h',[mom(1),mom(2),mom(3)],SS.tr);
|
||||
# clf
|
||||
# ind=find(Ac>Hs/2);
|
||||
# plot(Tc(ind), Ac(ind),'.');
|
||||
# hold on
|
||||
# pdfplot(flh_g,'k-.')
|
||||
# pdfplot(f_tcac_s)
|
||||
# toc
|
||||
# wafostamp([],'(ER)')
|
||||
# disp('Block = 39'), pause(pstate)
|
||||
#
|
||||
# #!#!
|
||||
# clf
|
||||
# #! f_tcac = spec2thpdf(SS,[],'TcAc',[0 12 81],[0:0.2:8],opt1);
|
||||
# #! pdfplot(f_tcac)
|
||||
# disp('Block = 40'), pause(pstate)
|
||||
#
|
||||
# #!#! Section 3.4.4 Joint density of crest and trough height
|
||||
# #!#! Section 3.4.5 Min-to-max distributions Markov method
|
||||
# #!#! Example 11. (min-max problems with Gullfaks data)
|
||||
# #!#! Joint density of maximum and the following minimum
|
||||
# clf
|
||||
# tic
|
||||
# tp = dat2tp(yy);
|
||||
# Mm = fliplr(tp2mm(tp));
|
||||
# fmm = kde(Mm);
|
||||
# f_mM = spec2mmtpdf(SS,[],'mm',[],[-7 7 51],opt2);
|
||||
#
|
||||
# pdfplot(f_mM,'-.')
|
||||
# hold on
|
||||
# pdfplot(fmm,'k-')
|
||||
# hold off
|
||||
# wafostamp([],'(ER)')
|
||||
# toc
|
||||
# disp('Block = 41'), pause(pstate)
|
||||
#
|
||||
# #!#! The joint density of still water separated maxima and minima.
|
||||
# clf
|
||||
# tic
|
||||
# ind = find(Mm(:,1)>v & Mm(:,2)<v);
|
||||
# Mmv = abs(Mm(ind,:)-v);
|
||||
# fmmv = kde(Mmv);
|
||||
# f_vmm = spec2mmtpdf(SS,[],'vmm',[],[-7 7 51],opt2);
|
||||
# clf
|
||||
# pdfplot(fmmv,'k-')
|
||||
# hold on
|
||||
# pdfplot(f_vmm,'-.')
|
||||
# hold off
|
||||
# wafostamp([],'(ER)')
|
||||
# toc
|
||||
# disp('Block = 42'), pause(pstate)
|
||||
#
|
||||
#
|
||||
# #!#!
|
||||
# clf
|
||||
# tic
|
||||
# facat = kde([Ac At]);
|
||||
# f_acat = spec2mmtpdf(SS,[],'AcAt',[],[-7 7 51],opt2);
|
||||
# clf
|
||||
# pdfplot(f_acat,'-.')
|
||||
# hold on
|
||||
# pdfplot(facat,'k-')
|
||||
# hold off
|
||||
# wafostamp([],'(ER)')
|
||||
# toc
|
||||
# disp('Block = 43'), pause(pstate)
|
||||
|
@ -1,407 +0,0 @@
|
||||
|
||||
#! CHAPTER4 contains the commands used in Chapter 4 of the tutorial
|
||||
#!=================================================================
|
||||
#!
|
||||
#! CALL: Chapter4
|
||||
#!
|
||||
#! Some of the commands are edited for fast computation.
|
||||
#! Each set of commands is followed by a 'pause' command.
|
||||
#!
|
||||
#! This routine also can print the figures;
|
||||
#! For printing the figures on directory ../bilder/ edit the file and put
|
||||
#! printing=1;
|
||||
|
||||
#! Tested on Matlab 5.3
|
||||
#! History
|
||||
#! Revised pab sept2005
|
||||
#! Added sections -> easier to evaluate using cellmode evaluation.
|
||||
#! revised pab Feb2004
|
||||
#! updated call to lc2sdat
|
||||
#! Created by GL July 13, 2000
|
||||
#! from commands used in Chapter 4
|
||||
#!
|
||||
|
||||
#! Chapter 4 Fatigue load analysis and rain-flow cycles
|
||||
#!------------------------------------------------------
|
||||
|
||||
printing = 0
|
||||
|
||||
|
||||
#! Section 4.3.1 Crossing intensity
|
||||
#!~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
|
||||
import numpy as np
|
||||
from wafo.plotbackend import plotbackend as plt
|
||||
import wafo.data as wd
|
||||
import wafo.objects as wo
|
||||
|
||||
xx_sea = wd.sea()
|
||||
ts = wo.mat2timeseries(xx_sea)
|
||||
tp = ts.turning_points()
|
||||
mM = tp.cycle_pairs(kind='min2max')
|
||||
lc = mM.level_crossings(intensity=True)
|
||||
T_sea = ts.args[-1]-ts.args[0]
|
||||
|
||||
plt.subplot(1,2,1)
|
||||
lc.plot()
|
||||
plt.subplot(1,2,2)
|
||||
lc.setplotter(plotmethod='step')
|
||||
lc.plot()
|
||||
plt.show()
|
||||
|
||||
|
||||
m_sea = ts.data.mean()
|
||||
f0_sea = np.interp(m_sea, lc.args,lc.data)
|
||||
extr_sea = len(tp.data)/(2*T_sea)
|
||||
alfa_sea = f0_sea/extr_sea
|
||||
print('alfa = %g ' % alfa_sea)
|
||||
|
||||
#! Section 4.3.2 Extraction of rainflow cycles
|
||||
#!~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
|
||||
#! Min-max and rainflow cycle plots
|
||||
#!~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
|
||||
mM_rfc = tp.cycle_pairs(h=0.3)
|
||||
|
||||
plt.clf()
|
||||
plt.subplot(122),
|
||||
mM.plot()
|
||||
plt.title('min-max cycle pairs')
|
||||
plt.subplot(121),
|
||||
mM_rfc.plot()
|
||||
plt.title('Rainflow filtered cycles')
|
||||
plt.show()
|
||||
|
||||
#! Min-max and rainflow cycle distributions
|
||||
#!~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
|
||||
import wafo.misc as wm
|
||||
ampmM_sea = mM.amplitudes()
|
||||
ampRFC_sea = mM_rfc.amplitudes()
|
||||
plt.clf()
|
||||
plt.subplot(121)
|
||||
wm.plot_histgrm(ampmM_sea,25)
|
||||
ylim = plt.gca().get_ylim()
|
||||
plt.title('min-max amplitude distribution')
|
||||
plt.subplot(122)
|
||||
wm.plot_histgrm(ampRFC_sea,25)
|
||||
plt.gca().set_ylim(ylim)
|
||||
plt.title('Rainflow amplitude distribution')
|
||||
plt.show()
|
||||
|
||||
#!#! Section 4.3.3 Simulation of rainflow cycles
|
||||
#!#! Simulation of cycles in a Markov model
|
||||
# n = 41
|
||||
# param_m = [-1, 1, n]
|
||||
# param_D = [1, n, n]
|
||||
# u_markov=levels(param_m);
|
||||
# G_markov=mktestmat(param_m,[-0.2, 0.2],0.15,1);
|
||||
# T_markov=5000;
|
||||
#xxD_markov=mctpsim({G_markov [,]},T_markov);
|
||||
#xx_markov=[(1:T_markov)' u_markov(xxD_markov)'];
|
||||
#clf
|
||||
#plot(xx_markov(1:50,1),xx_markov(1:50,2))
|
||||
#title('Markov chain of turning points')
|
||||
#wafostamp([],'(ER)')
|
||||
#disp('Block 5'),pause(pstate)
|
||||
#
|
||||
#
|
||||
##!#! Rainflow cycles in a transformed Gaussian model
|
||||
##!#! Hermite transformed wave data and rainflow filtered turning points, h = 0.2.
|
||||
#me = mean(xx_sea(:,2));
|
||||
#sa = std(xx_sea(:,2));
|
||||
#Hm0_sea = 4*sa;
|
||||
#Tp_sea = 1/max(lc_sea(:,2));
|
||||
#spec = jonswap([],[Hm0_sea Tp_sea]);
|
||||
#
|
||||
#[sk, ku] = spec2skew(spec);
|
||||
#spec.tr = hermitetr([],[sa sk ku me]);
|
||||
#param_h = [-1.5 2 51];
|
||||
#spec_norm = spec;
|
||||
#spec_norm.S = spec_norm.S/sa^2;
|
||||
#xx_herm = spec2sdat(spec_norm,[2^15 1],0.1);
|
||||
##! ????? PJ, JR 11-Apr-2001
|
||||
##! NOTE, in the simulation program spec2sdat
|
||||
##!the spectrum must be normalized to variance 1
|
||||
##! ?????
|
||||
#h = 0.2;
|
||||
#[dtp,u_herm,xx_herm_1]=dat2dtp(param_h,xx_herm,h);
|
||||
#clf
|
||||
#plot(xx_herm(:,1),xx_herm(:,2),'k','LineWidth',2); hold on;
|
||||
#plot(xx_herm_1(:,1),xx_herm_1(:,2),'k--','Linewidth',2);
|
||||
#axis([0 50 -1 1]), hold off;
|
||||
#title('Rainflow filtered wave data')
|
||||
#wafostamp([],'(ER)')
|
||||
#disp('Block 6'),pause(pstate)
|
||||
#
|
||||
##!#! Rainflow cycles and rainflow filtered rainflow cycles in the transformed Gaussian process.
|
||||
#tp_herm=dat2tp(xx_herm);
|
||||
#RFC_herm=tp2rfc(tp_herm);
|
||||
#mM_herm=tp2mm(tp_herm);
|
||||
#h=0.2;
|
||||
#[dtp,u,tp_herm_1]=dat2dtp(param_h,xx_herm,h);
|
||||
#RFC_herm_1 = tp2rfc(tp_herm_1);
|
||||
#clf
|
||||
#subplot(121), ccplot(RFC_herm)
|
||||
#title('h=0')
|
||||
#subplot(122), ccplot(RFC_herm_1)
|
||||
#title('h=0.2')
|
||||
#if (printing==1), print -deps ../bilder/fatigue_8.eps
|
||||
#end
|
||||
#wafostamp([],'(ER)')
|
||||
#disp('Block 7'),pause(pstate)
|
||||
#
|
||||
##!#! Section 4.3.4 Calculating the rainflow matrix
|
||||
#
|
||||
#
|
||||
#Grfc_markov=mctp2rfm({G_markov []});
|
||||
#clf
|
||||
#subplot(121), cmatplot(u_markov,u_markov,G_markov), axis('square')
|
||||
#subplot(122), cmatplot(u_markov,u_markov,Grfc_markov), axis('square')
|
||||
#wafostamp([],'(ER)')
|
||||
#disp('Block 8'),pause(pstate)
|
||||
#
|
||||
##!#!
|
||||
#clf
|
||||
#cmatplot(u_markov,u_markov,{G_markov Grfc_markov},3)
|
||||
#wafostamp([],'(ER)')
|
||||
#disp('Block 9'),pause(pstate)
|
||||
#
|
||||
##!#! Min-max-matrix and theoretical rainflow matrix for test Markov sequence.
|
||||
#cmatplot(u_markov,u_markov,{G_markov Grfc_markov},4)
|
||||
#subplot(121), axis('square'), title('min2max transition matrix')
|
||||
#subplot(122), axis('square'), title('Rainflow matrix')
|
||||
#if (printing==1), print -deps ../bilder/fatigue_9.eps
|
||||
#end
|
||||
#wafostamp([],'(ER)')
|
||||
#disp('Block 10'),pause(pstate)
|
||||
#
|
||||
##!#! Observed and theoretical rainflow matrix for test Markov sequence.
|
||||
#n=length(u_markov);
|
||||
#Frfc_markov=dtp2rfm(xxD_markov,n);
|
||||
#clf
|
||||
#cmatplot(u_markov,u_markov,{Frfc_markov Grfc_markov*T_markov/2},3)
|
||||
#subplot(121), axis('square'), title('Observed rainflow matrix')
|
||||
#subplot(122), axis('square'), title('Theoretical rainflow matrix')
|
||||
#if (printing==1), print -deps ../bilder/fatigue_10.eps
|
||||
#end
|
||||
#wafostamp([],'(ER)')
|
||||
#disp('Block 11'),pause(pstate)
|
||||
#
|
||||
##!#! Smoothed observed and calculated rainflow matrix for test Markov sequence.
|
||||
#tp_markov=dat2tp(xx_markov);
|
||||
#RFC_markov=tp2rfc(tp_markov);
|
||||
#h=1;
|
||||
#Frfc_markov_smooth=cc2cmat(param_m,RFC_markov,[],1,h);
|
||||
#clf
|
||||
#cmatplot(u_markov,u_markov,{Frfc_markov_smooth Grfc_markov*T_markov/2},4)
|
||||
#subplot(121), axis('square'), title('Smoothed observed rainflow matrix')
|
||||
#subplot(122), axis('square'), title('Theoretical rainflow matrix')
|
||||
#if (printing==1), print -deps ../bilder/fatigue_11.eps
|
||||
#end
|
||||
#wafostamp([],'(ER)')
|
||||
#disp('Block 12'),pause(pstate)
|
||||
#
|
||||
##!#! Rainflow matrix from spectrum
|
||||
#clf
|
||||
##!GmM3_herm=spec2mmtpdf(spec,[],'Mm',[],[],2);
|
||||
#GmM3_herm=spec2cmat(spec,[],'Mm',[],param_h,2);
|
||||
#pdfplot(GmM3_herm)
|
||||
#wafostamp([],'(ER)')
|
||||
#disp('Block 13'),pause(pstate)
|
||||
#
|
||||
#
|
||||
##!#! Min-max matrix and theoretical rainflow matrix for Hermite-transformed Gaussian waves.
|
||||
#Grfc_herm=mctp2rfm({GmM3_herm.f []});
|
||||
#u_herm=levels(param_h);
|
||||
#clf
|
||||
#cmatplot(u_herm,u_herm,{GmM3_herm.f Grfc_herm},4)
|
||||
#subplot(121), axis('square'), title('min-max matrix')
|
||||
#subplot(122), axis('square'), title('Theoretical rainflow matrix')
|
||||
#if (printing==1), print -deps ../bilder/fatigue_12.eps
|
||||
#end
|
||||
#wafostamp([],'(ER)')
|
||||
#disp('Block 14'),pause(pstate)
|
||||
#
|
||||
##!#!
|
||||
#clf
|
||||
#Grfc_direct_herm=spec2cmat(spec,[],'rfc',[],[],2);
|
||||
#subplot(121), pdfplot(GmM3_herm), axis('square'), hold on
|
||||
#subplot(122), pdfplot(Grfc_direct_herm), axis('square'), hold off
|
||||
#if (printing==1), print -deps ../bilder/fig_mmrfcjfr.eps
|
||||
#end
|
||||
#wafostamp([],'(ER)')
|
||||
#disp('Block 15'),pause(pstate)
|
||||
#
|
||||
#
|
||||
##!#! Observed smoothed and theoretical min-max matrix,
|
||||
##!#! (and observed smoothed and theoretical rainflow matrix for Hermite-transformed Gaussian waves).
|
||||
#tp_herm=dat2tp(xx_herm);
|
||||
#RFC_herm=tp2rfc(tp_herm);
|
||||
#mM_herm=tp2mm(tp_herm);
|
||||
#h=0.2;
|
||||
#FmM_herm_smooth=cc2cmat(param_h,mM_herm,[],1,h);
|
||||
#Frfc_herm_smooth=cc2cmat(param_h,RFC_herm,[],1,h);
|
||||
#T_herm=xx_herm(end,1)-xx_herm(1,1);
|
||||
#clf
|
||||
#cmatplot(u_herm,u_herm,{FmM_herm_smooth GmM3_herm.f*length(mM_herm) ; ...
|
||||
# Frfc_herm_smooth Grfc_herm*length(RFC_herm)},4)
|
||||
#subplot(221), axis('square'), title('Observed smoothed min-max matrix')
|
||||
#subplot(222), axis('square'), title('Theoretical min-max matrix')
|
||||
#subplot(223), axis('square'), title('Observed smoothed rainflow matrix')
|
||||
#subplot(224), axis('square'), title('Theoretical rainflow matrix')
|
||||
#if (printing==1), print -deps ../bilder/fatigue_13.eps
|
||||
#end
|
||||
#wafostamp([],'(ER)')
|
||||
#disp('Block 16'),pause(pstate)
|
||||
#
|
||||
##!#! Section 4.3.5 Simulation from crossings and rainflow structure
|
||||
#
|
||||
##!#! Crossing spectrum (smooth curve) and obtained spectrum (wiggled curve)
|
||||
##!#! for simulated process with irregularity factor 0.25.
|
||||
#clf
|
||||
#cross_herm=dat2lc(xx_herm);
|
||||
#alpha1=0.25;
|
||||
#alpha2=0.75;
|
||||
#xx_herm_sim1=lc2sdat(cross_herm,500,alpha1);
|
||||
#cross_herm_sim1=dat2lc(xx_herm_sim1);
|
||||
#subplot(211)
|
||||
#plot(cross_herm(:,1),cross_herm(:,2)/max(cross_herm(:,2)))
|
||||
#hold on
|
||||
#stairs(cross_herm_sim1(:,1),...
|
||||
# cross_herm_sim1(:,2)/max(cross_herm_sim1(:,2)))
|
||||
#hold off
|
||||
#title('Crossing intensity, \alpha = 0.25')
|
||||
#subplot(212)
|
||||
#plot(xx_herm_sim1(:,1),xx_herm_sim1(:,2))
|
||||
#title('Simulated load, \alpha = 0.25')
|
||||
#if (printing==1), print -deps ../bilder/fatigue_14_25.eps
|
||||
#end
|
||||
#wafostamp([],'(ER)')
|
||||
#disp('Block 16'),pause(pstate)
|
||||
#
|
||||
##!#! Crossing spectrum (smooth curve) and obtained spectrum (wiggled curve)
|
||||
##!#! for simulated process with irregularity factor 0.75.
|
||||
#xx_herm_sim2=lc2sdat(cross_herm,500,alpha2);
|
||||
#cross_herm_sim2=dat2lc(xx_herm_sim2);
|
||||
#subplot(211)
|
||||
#plot(cross_herm(:,1),cross_herm(:,2)/max(cross_herm(:,2)))
|
||||
#hold on
|
||||
#stairs(cross_herm_sim2(:,1),...
|
||||
# cross_herm_sim2(:,2)/max(cross_herm_sim2(:,2)))
|
||||
#hold off
|
||||
#title('Crossing intensity, \alpha = 0.75')
|
||||
#subplot(212)
|
||||
#plot(xx_herm_sim2(:,1),xx_herm_sim2(:,2))
|
||||
#title('Simulated load, \alpha = 0.75')
|
||||
#if (printing==1), print -deps ../bilder/fatigue_14_75.eps
|
||||
#end
|
||||
#wafostamp([],'(ER)')
|
||||
#disp('Block 17'),pause(pstate)
|
||||
#
|
||||
##!#! Section 4.4 Fatigue damage and fatigue life distribution
|
||||
##!#! Section 4.4.1 Introduction
|
||||
#beta=3.2; gam=5.5E-10; T_sea=xx_sea(end,1)-xx_sea(1,1);
|
||||
#d_beta=cc2dam(RFC_sea,beta)/T_sea;
|
||||
#time_fail=1/gam/d_beta/3600 #!in hours of the specific storm
|
||||
#disp('Block 18'),pause(pstate)
|
||||
#
|
||||
##!#! Section 4.4.2 Level crossings
|
||||
##!#! Crossing intensity as calculated from the Markov matrix (solid curve) and from the observed rainflow matrix (dashed curve).
|
||||
#clf
|
||||
#mu_markov=cmat2lc(param_m,Grfc_markov);
|
||||
#muObs_markov=cmat2lc(param_m,Frfc_markov/(T_markov/2));
|
||||
#clf
|
||||
#plot(mu_markov(:,1),mu_markov(:,2),muObs_markov(:,1),muObs_markov(:,2),'--')
|
||||
#title('Theoretical and observed crossing intensity ')
|
||||
#if (printing==1), print -deps ../bilder/fatigue_15.eps
|
||||
#end
|
||||
#wafostamp([],'(ER)')
|
||||
#disp('Block 19'),pause(pstate)
|
||||
#
|
||||
##!#! Section 4.4.3 Damage
|
||||
##!#! Distribution of damage from different RFC cycles, from calculated theoretical and from observed rainflow matrix.
|
||||
#beta = 4;
|
||||
#Dam_markov = cmat2dam(param_m,Grfc_markov,beta)
|
||||
#DamObs1_markov = cc2dam(RFC_markov,beta)/(T_markov/2)
|
||||
#DamObs2_markov = cmat2dam(param_m,Frfc_markov,beta)/(T_markov/2)
|
||||
#disp('Block 20'),pause(pstate)
|
||||
#
|
||||
#Dmat_markov = cmat2dmat(param_m,Grfc_markov,beta);
|
||||
#DmatObs_markov = cmat2dmat(param_m,Frfc_markov,beta)/(T_markov/2);
|
||||
#clf
|
||||
#subplot(121), cmatplot(u_markov,u_markov,Dmat_markov,4)
|
||||
#title('Theoretical damage matrix')
|
||||
#subplot(122), cmatplot(u_markov,u_markov,DmatObs_markov,4)
|
||||
#title('Observed damage matrix')
|
||||
#if (printing==1), print -deps ../bilder/fatigue_16.eps
|
||||
#end
|
||||
#wafostamp([],'(ER)')
|
||||
#disp('Block 21'),pause(pstate)
|
||||
#
|
||||
#
|
||||
##!#!
|
||||
##!Damplus_markov = lc2dplus(mu_markov,beta)
|
||||
#pause(pstate)
|
||||
#
|
||||
##!#! Section 4.4.4 Estimation of S-N curve
|
||||
#
|
||||
##!#! Load SN-data and plot in log-log scale.
|
||||
#SN = load('sn.dat');
|
||||
#s = SN(:,1);
|
||||
#N = SN(:,2);
|
||||
#clf
|
||||
#loglog(N,s,'o'), axis([0 14e5 10 30])
|
||||
##!if (printing==1), print -deps ../bilder/fatigue_?.eps end
|
||||
#wafostamp([],'(ER)')
|
||||
#disp('Block 22'),pause(pstate)
|
||||
#
|
||||
#
|
||||
##!#! Check of S-N-model on normal probability paper.
|
||||
#
|
||||
#normplot(reshape(log(N),8,5))
|
||||
#if (printing==1), print -deps ../bilder/fatigue_17.eps
|
||||
#end
|
||||
#wafostamp([],'(ER)')
|
||||
#disp('Block 23'),pause(pstate)
|
||||
#
|
||||
##!#! Estimation of S-N-model on linear scale.
|
||||
#clf
|
||||
#[e0,beta0,s20] = snplot(s,N,12);
|
||||
#title('S-N-data with estimated N(s)','FontSize',20)
|
||||
#set(gca,'FontSize',20)
|
||||
#if (printing==1), print -deps ../bilder/fatigue_18a.eps
|
||||
#end
|
||||
#wafostamp([],'(ER)')
|
||||
#disp('Block 24'),pause(pstate)
|
||||
#
|
||||
##!#! Estimation of S-N-model on log-log scale.
|
||||
#clf
|
||||
#[e0,beta0,s20] = snplot(s,N,14);
|
||||
#title('S-N-data with estimated N(s)','FontSize',20)
|
||||
#set(gca,'FontSize',20)
|
||||
#if (printing==1), print -deps ../bilder/fatigue_18b.eps
|
||||
#end
|
||||
#wafostamp([],'(ER)')
|
||||
#disp('Block 25'),pause(pstate)
|
||||
#
|
||||
##!#! Section 4.4.5 From S-N curve to fatigue life distribution
|
||||
##!#! Damage intensity as function of $\beta$
|
||||
#beta = 3:0.1:8;
|
||||
#DRFC = cc2dam(RFC_sea,beta);
|
||||
#dRFC = DRFC/T_sea;
|
||||
#plot(beta,dRFC), axis([3 8 0 0.25])
|
||||
#title('Damage intensity as function of \beta')
|
||||
#if (printing==1), print -deps ../bilder/fatigue_19.eps
|
||||
#end
|
||||
#wafostamp([],'(ER)')
|
||||
#disp('Block 26'),pause(pstate)
|
||||
#
|
||||
##!#! Fatigue life distribution with sea load.
|
||||
#dam0 = cc2dam(RFC_sea,beta0)/T_sea;
|
||||
#[t0,F0] = ftf(e0,dam0,s20,0.5,1);
|
||||
#[t1,F1] = ftf(e0,dam0,s20,0,1);
|
||||
#[t2,F2] = ftf(e0,dam0,s20,5,1);
|
||||
#plot(t0,F0,t1,F1,t2,F2)
|
||||
#title('Fatigue life distribution function')
|
||||
#if (printing==1), print -deps ../bilder/fatigue_20.eps
|
||||
#end
|
||||
#wafostamp([],'(ER)')
|
||||
#disp('Block 27, last block')
|
@ -1,238 +0,0 @@
|
||||
## CHAPTER5 contains the commands used in Chapter 5 of the tutorial
|
||||
#
|
||||
# CALL: Chapter5
|
||||
#
|
||||
# Some of the commands are edited for fast computation.
|
||||
# Each set of commands is followed by a 'pause' command.
|
||||
#
|
||||
|
||||
# Tested on Matlab 5.3
|
||||
# History
|
||||
# Added Return values by GL August 2008
|
||||
# Revised pab sept2005
|
||||
# Added sections -> easier to evaluate using cellmode evaluation.
|
||||
# Created by GL July 13, 2000
|
||||
# from commands used in Chapter 5
|
||||
#
|
||||
|
||||
## Chapter 5 Extreme value analysis
|
||||
|
||||
## Section 5.1 Weibull and Gumbel papers
|
||||
from __future__ import division
|
||||
import numpy as np
|
||||
import scipy.interpolate as si
|
||||
from wafo.plotbackend import plotbackend as plt
|
||||
import wafo.data as wd
|
||||
import wafo.objects as wo
|
||||
import wafo.stats as ws
|
||||
import wafo.kdetools as wk
|
||||
pstate = 'off'
|
||||
|
||||
# Significant wave-height data on Weibull paper,
|
||||
|
||||
fig = plt.figure()
|
||||
ax = fig.add_subplot(111)
|
||||
Hs = wd.atlantic()
|
||||
wei = ws.weibull_min.fit(Hs)
|
||||
tmp = ws.probplot(Hs, wei, ws.weibull_min, plot=ax)
|
||||
plt.show()
|
||||
#wafostamp([],'(ER)')
|
||||
#disp('Block = 1'),pause(pstate)
|
||||
|
||||
##
|
||||
# Significant wave-height data on Gumbel paper,
|
||||
plt.clf()
|
||||
ax = fig.add_subplot(111)
|
||||
gum = ws.gumbel_r.fit(Hs)
|
||||
tmp1 = ws.probplot(Hs, gum, ws.gumbel_r, plot=ax)
|
||||
#wafostamp([],'(ER)')
|
||||
plt.show()
|
||||
#disp('Block = 2'),pause(pstate)
|
||||
|
||||
##
|
||||
# Significant wave-height data on Normal probability paper,
|
||||
plt.clf()
|
||||
ax = fig.add_subplot(111)
|
||||
phat = ws.norm.fit2(np.log(Hs))
|
||||
phat.plotresq()
|
||||
#tmp2 = ws.probplot(np.log(Hs), phat, ws.norm, plot=ax)
|
||||
|
||||
#wafostamp([],'(ER)')
|
||||
plt.show()
|
||||
#disp('Block = 3'),pause(pstate)
|
||||
|
||||
##
|
||||
# Return values in the Gumbel distribution
|
||||
plt.clf()
|
||||
T = np.r_[1:100000]
|
||||
sT = gum[0] - gum[1] * np.log(-np.log1p(-1./T))
|
||||
plt.semilogx(T, sT)
|
||||
plt.hold(True)
|
||||
# ws.edf(Hs).plot()
|
||||
Nmax = len(Hs)
|
||||
N = np.r_[1:Nmax + 1]
|
||||
|
||||
plt.plot(Nmax/N, sorted(Hs, reverse=True), '.')
|
||||
plt.title('Return values in the Gumbel model')
|
||||
plt.xlabel('Return period')
|
||||
plt.ylabel('Return value')
|
||||
#wafostamp([],'(ER)')
|
||||
plt.show()
|
||||
#disp('Block = 4'),pause(pstate)
|
||||
|
||||
## Section 5.2 Generalized Pareto and Extreme Value distributions
|
||||
## Section 5.2.1 Generalized Extreme Value distribution
|
||||
|
||||
# Empirical distribution of significant wave-height with estimated
|
||||
# Generalized Extreme Value distribution,
|
||||
gev = ws.genextreme.fit2(Hs)
|
||||
gev.plotfitsummary()
|
||||
# wafostamp([],'(ER)')
|
||||
# disp('Block = 5a'),pause(pstate)
|
||||
|
||||
plt.clf()
|
||||
x = np.linspace(0,14,200)
|
||||
kde = wk.TKDE(Hs, L2=0.5)(x, output='plot')
|
||||
kde.plot()
|
||||
plt.hold(True)
|
||||
plt.plot(x, gev.pdf(x),'--')
|
||||
# disp('Block = 5b'),pause(pstate)
|
||||
|
||||
# Analysis of yura87 wave data.
|
||||
# Wave data interpolated (spline) and organized in 5-minute intervals
|
||||
# Normalized to mean 0 and std = 1 to get stationary conditions.
|
||||
# maximum level over each 5-minute interval analysed by GEV
|
||||
xn = wd.yura87()
|
||||
XI = np.r_[1:len(xn):0.25] - .99
|
||||
N = len(XI)
|
||||
N = N - np.mod(N, 4*60*5)
|
||||
|
||||
YI = si.interp1d(xn[:, 0], xn[:, 1], kind='linear')(XI)
|
||||
YI = YI.reshape(4*60*5, N/(4*60*5)) # Each column holds 5 minutes of
|
||||
# interpolated data.
|
||||
Y5 = (YI - YI.mean(axis=0)) / YI.std(axis=0)
|
||||
Y5M = Y5.maximum(axis=0)
|
||||
Y5gev = ws.genextreme.fit2(Y5M,method='mps')
|
||||
Y5gev.plotfitsummary()
|
||||
#wafostamp([],'(ER)')
|
||||
#disp('Block = 6'),pause(pstate)
|
||||
|
||||
## Section 5.2.2 Generalized Pareto distribution
|
||||
|
||||
# Exceedances of significant wave-height data over level 3,
|
||||
gpd3 = ws.genpareto.fit2(Hs[Hs>3]-3, floc=0)
|
||||
gpd3.plotfitsummary()
|
||||
#wafostamp([],'(ER)')
|
||||
|
||||
##
|
||||
plt.figure()
|
||||
# Exceedances of significant wave-height data over level 7,
|
||||
gpd7 = ws.genpareto.fit2(Hs(Hs>7), floc=7)
|
||||
gpd7.plotfitsummary()
|
||||
# wafostamp([],'(ER)')
|
||||
# disp('Block = 6'),pause(pstate)
|
||||
|
||||
##
|
||||
#Simulates 100 values from the GEV distribution with parameters (0.3, 1, 2),
|
||||
# then estimates the parameters using two different methods and plots the
|
||||
# estimated distribution functions together with the empirical distribution.
|
||||
Rgev = ws.genextreme.rvs(0.3,1,2,size=100)
|
||||
gp = ws.genextreme.fit2(Rgev, method='mps');
|
||||
gm = ws.genextreme.fit2(Rgev, *gp.par.tolist(), method='ml')
|
||||
gm.plotfitsummary()
|
||||
|
||||
gp.plotecdf()
|
||||
plt.hold(True)
|
||||
plt.plot(x, gm.cdf(x), '--')
|
||||
plt.hold(False)
|
||||
#wafostamp([],'(ER)')
|
||||
#disp('Block =7'),pause(pstate)
|
||||
|
||||
##
|
||||
# ;
|
||||
Rgpd = ws.genpareto.rvs(0.4,0, 1,size=100)
|
||||
gp = ws.genpareto.fit2(Rgpd, method='mps')
|
||||
gml = ws.genpareto.fit2(Rgpd, method='ml')
|
||||
|
||||
gp.plotecdf()
|
||||
x = sorted(Rgpd)
|
||||
plt.hold(True)
|
||||
plt.plot(x, gml.cdf(x))
|
||||
# gm = fitgenpar(Rgpd,'method','mom','plotflag',0);
|
||||
# plot(x,cdfgenpar(x,gm),'g--')
|
||||
#gw = fitgenpar(Rgpd,'method','pwm','plotflag',0);
|
||||
#plot(x,cdfgenpar(x,gw),'g:')
|
||||
#gml = fitgenpar(Rgpd,'method','ml','plotflag',0);
|
||||
#plot(x,cdfgenpar(x,gml),'--')
|
||||
#gmps = fitgenpar(Rgpd,'method','mps','plotflag',0);
|
||||
#plot(x,cdfgenpar(x,gmps),'r-.')
|
||||
plt.hold(False)
|
||||
#wafostamp([],'(ER)')
|
||||
#disp('Block = 8'),pause(pstate)
|
||||
|
||||
##
|
||||
# Return values for the GEV distribution
|
||||
T = np.logspace(1, 5, 10);
|
||||
#[sT, sTlo, sTup] = invgev(1./T,Y5gev,'lowertail',false,'proflog',true);
|
||||
|
||||
#T = 2:100000;
|
||||
#k=Y5gev.params(1); mu=Y5gev.params(3); sigma=Y5gev.params(2);
|
||||
#sT1 = invgev(1./T,Y5gev,'lowertail',false);
|
||||
#sT=mu + sigma/k*(1-(-log(1-1./T)).^k);
|
||||
plt.clf()
|
||||
#plt.semilogx(T,sT,T,sTlo,'r',T,sTup,'r')
|
||||
#plt.hold(True)
|
||||
#N = np.r_[1:len(Y5M)]
|
||||
#Nmax = max(N);
|
||||
#plot(Nmax./N, sorted(Y5M,reverse=True), '.')
|
||||
#plt.title('Return values in the GEV model')
|
||||
#plt.xlabel('Return priod')
|
||||
#plt.ylabel('Return value')
|
||||
#plt.grid(True)
|
||||
#disp('Block = 9'),pause(pstate)
|
||||
|
||||
## Section 5.3 POT-analysis
|
||||
|
||||
# Estimated expected exceedance over level u as function of u.
|
||||
plt.clf()
|
||||
|
||||
mrl = ws.reslife(Hs,'umin',2,'umax',10,'Nu',200);
|
||||
mrl.plot()
|
||||
#wafostamp([],'(ER)')
|
||||
#disp('Block = 10'),pause(pstate)
|
||||
|
||||
##
|
||||
# Estimated distribution functions of monthly maxima
|
||||
#with the POT method (solid),
|
||||
# fitting a GEV (dashed) and the empirical distribution.
|
||||
|
||||
# POT- method
|
||||
gpd7 = ws.genpareto.fit2(Hs(Hs>7)-7, method='mps', floc=0)
|
||||
khat, loc, sigmahat = gpd7.par
|
||||
|
||||
muhat = len(Hs[Hs>7])/(7*3*2)
|
||||
bhat = sigmahat/muhat**khat
|
||||
ahat = 7-(bhat-sigmahat)/khat
|
||||
x = np.linspace(5,15,200);
|
||||
plt.plot(x,ws.genextreme.cdf(x, khat,bhat,ahat))
|
||||
# disp('Block = 11'),pause(pstate)
|
||||
|
||||
##
|
||||
# Since we have data to compute the monthly maxima mm over
|
||||
#42 months we can also try to fit a
|
||||
# GEV distribution directly:
|
||||
mm = np.zeros((1,41))
|
||||
for i in range(41):
|
||||
mm[i] = max(Hs[((i-1)*14+1):i*14])
|
||||
|
||||
|
||||
gev = ws.genextreme.fit2(mm)
|
||||
|
||||
|
||||
plt.hold(True)
|
||||
gev.plotecdf()
|
||||
|
||||
plt.hold(False)
|
||||
#wafostamp([],'(ER)')
|
||||
#disp('Block = 12, Last block'),pause(pstate)
|
||||
|
@ -1,176 +0,0 @@
|
||||
from tutor_init import *
|
||||
import itertools
|
||||
# import sys
|
||||
log = logging.getLogger(__name__)
|
||||
log.setLevel(logging.DEBUG)
|
||||
|
||||
MARKERS = ('o', 'x', '+', '.', '<', '>', '^', 'v')
|
||||
|
||||
|
||||
def plot_varying_symbols(x, y, color='red', size=5):
|
||||
"""
|
||||
Create a plot with varying symbols
|
||||
Parameters
|
||||
----------
|
||||
x : numpy array with x data of the points
|
||||
y : numpy array with y data of the points
|
||||
color : color of the symbols
|
||||
|
||||
Returns
|
||||
-------
|
||||
|
||||
"""
|
||||
markers = itertools.cycle(MARKERS)
|
||||
for q, p in zip(x, y):
|
||||
plt.plot(q, p, marker=markers.next(), linestyle='', color=color,
|
||||
markersize=size)
|
||||
|
||||
|
||||
def damage_vs_S(S, beta, K):
|
||||
"""
|
||||
Calculate the damage 1/N for a given stress S
|
||||
|
||||
Parameters
|
||||
----------
|
||||
S : Stress [Pa]
|
||||
beta : coefficient, typically 3
|
||||
K : constant
|
||||
|
||||
Returns
|
||||
-------
|
||||
|
||||
"""
|
||||
return K * np.power(S, beta)
|
||||
|
||||
# Section 4.3.1 Crossing intensity
|
||||
# ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
|
||||
import wafo.data as wd
|
||||
import wafo.objects as wo
|
||||
import wafo.misc as wm
|
||||
|
||||
xx_sea = wd.sea()
|
||||
|
||||
Tlength = xx_sea[-1, 0] - xx_sea[0, 0]
|
||||
beta = 3
|
||||
K1 = 6.5e-31
|
||||
Np = 200
|
||||
Tp = Tlength / Np
|
||||
A = 100e6
|
||||
log.info("setting sin wave with Tp={} and T={}".format(Tp, Tlength))
|
||||
Nc = 1.0 / damage_vs_S(A, beta, K1)
|
||||
damage = float(Np) / float(Nc)
|
||||
log.info("budget at S={} N={}: damage = {} ".format(A, Nc, damage))
|
||||
#xx_sea[:, 1] = A * np.cos(2 * np.pi * xx_sea[:, 0]/Tp)
|
||||
xx_sea[:, 1] *= 500e6
|
||||
|
||||
log.info("loaded sea time series {}".format(xx_sea.shape))
|
||||
ts = wo.mat2timeseries(xx_sea)
|
||||
|
||||
tp = ts.turning_points()
|
||||
mM = tp.cycle_pairs(kind='min2max')
|
||||
Mm = tp.cycle_pairs(kind='max2min')
|
||||
lc = mM.level_crossings(intensity=True)
|
||||
T_sea = ts.args[-1] - ts.args[0]
|
||||
|
||||
# for i in dir(mM):
|
||||
# print(i)
|
||||
|
||||
|
||||
ts1 = wo.mat2timeseries(xx_sea[:, :])
|
||||
tp1 = ts1.turning_points()
|
||||
sig_tp = ts.turning_points(h=0, wavetype='astm')
|
||||
try:
|
||||
sig_cp = sig_tp.cycle_astm()
|
||||
log.info("Successfully used cycle_astm")
|
||||
except AttributeError:
|
||||
log.warning("Could use cycle_astm")
|
||||
sig_cp = None
|
||||
tp1 = ts1.turning_points()
|
||||
tp2 = ts1.turning_points(wavetype='Mw')
|
||||
mM1 = tp1.cycle_pairs(kind='min2max')
|
||||
Mm1 = tp1.cycle_pairs(kind='max2min')
|
||||
|
||||
tp_rfc = tp1.rainflow_filter(h=100e6)
|
||||
mM_rfc = tp_rfc.cycle_pairs()
|
||||
try:
|
||||
mM_rfc_a = tp1.cycle_astm()
|
||||
except AttributeError:
|
||||
mM_rfc_a = None
|
||||
tc1 = ts1.trough_crest()
|
||||
min_to_max = True
|
||||
rfc_plot = True
|
||||
if min_to_max:
|
||||
m1, M1 = mM1.get_minima_and_maxima()
|
||||
i_min_start = 0
|
||||
else:
|
||||
m1, M1 = Mm1.get_minima_and_maxima()
|
||||
i_min_start = 2
|
||||
|
||||
m_rfc, M_rfc = mM_rfc.get_minima_and_maxima()
|
||||
# m_rfc_a, M_rfc_a = mM_rfc_a.get_minima_and_maxima()
|
||||
ts1.plot('b-')
|
||||
if rfc_plot:
|
||||
plot_varying_symbols(tp_rfc.args[0::2], m_rfc, color='red', size=10)
|
||||
plot_varying_symbols(tp_rfc.args[1::2], M_rfc, color='green', size=10)
|
||||
else:
|
||||
plot_varying_symbols(tp.args[i_min_start::2], m1, color='red', size=10)
|
||||
plot_varying_symbols(tp.args[1::2], M1, color='green', size=10)
|
||||
|
||||
set_windows_title("Sea time series", log)
|
||||
|
||||
plt.figure()
|
||||
plt.subplot(122),
|
||||
mM.plot()
|
||||
plt.title('min-max cycle pairs')
|
||||
plt.subplot(121),
|
||||
mM_rfc.plot()
|
||||
|
||||
title = 'Rainflow filtered cycles'
|
||||
plt.title(title)
|
||||
set_windows_title(title)
|
||||
|
||||
|
||||
# Min-max and rainflow cycle distributions
|
||||
#~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
|
||||
# import wafo.misc as wm
|
||||
ampmM_sea = mM.amplitudes()
|
||||
ampRFC_sea = mM_rfc.amplitudes()
|
||||
plt.figure()
|
||||
title = "s_n_curve"
|
||||
set_windows_title(title)
|
||||
S = np.linspace(1e6, 1000e6)
|
||||
plt.loglog(S, damage_vs_S(S, beta, K1))
|
||||
plt.figure()
|
||||
plt.subplot(121)
|
||||
stress_range = (1, 1e9)
|
||||
n_bins = 100
|
||||
wm.plot_histgrm(ampmM_sea, bins=n_bins, range=stress_range)
|
||||
plt.xlim(stress_range)
|
||||
ylim = plt.gca().get_ylim()
|
||||
plt.title('min-max amplitude distribution')
|
||||
plt.subplot(122)
|
||||
if sig_cp is not None:
|
||||
wm.plot_histgrm(sig_cp[:, 0], bins=n_bins, range=stress_range)
|
||||
plt.gca().set_ylim(ylim)
|
||||
title = 'Rainflow amplitude distribution'
|
||||
plt.title(title)
|
||||
plt.semilogy
|
||||
set_windows_title(title)
|
||||
|
||||
hist, bin_edges = np.histogram(
|
||||
sig_cp[
|
||||
:, 0], bins=n_bins, range=stress_range)
|
||||
|
||||
plt.figure()
|
||||
title = "my_bins"
|
||||
plt.title(title)
|
||||
plt.title(title)
|
||||
set_windows_title(title)
|
||||
plt.semilogy
|
||||
plt.bar(bin_edges[:-1], hist, width=stress_range[1] / n_bins)
|
||||
|
||||
print("damage min/max : {}".format(mM_rfc.damage([beta], K1)))
|
||||
|
||||
damage_rfc = K1 * np.sum(sig_cp[:, 0] ** beta)
|
||||
print("damage rfc : {}".format(damage_rfc))
|
||||
plt.show('hold')
|
@ -1,27 +0,0 @@
|
||||
import logging
|
||||
import scipy as sp
|
||||
import numpy as np
|
||||
from numpy import pi, reshape
|
||||
import matplotlib.pyplot as plt
|
||||
import matplotlib
|
||||
matplotlib.use('Qt4Agg')
|
||||
from matplotlib import rcParams
|
||||
rcParams.update({"font.size": 10})
|
||||
|
||||
|
||||
try:
|
||||
from win32api import LoadResource
|
||||
except ImportError:
|
||||
pass
|
||||
|
||||
log = logging.basicConfig(
|
||||
format="%(asctime)s - %(name)s - %(levelname)s - %(message)s",
|
||||
level=logging.DEBUG)
|
||||
|
||||
|
||||
def set_windows_title(title, log=None):
|
||||
if log is not None:
|
||||
log.info("Set windows title {}".format(title))
|
||||
fig = plt.gcf()
|
||||
fig.canvas.set_window_title(title)
|
||||
plt.show()
|
@ -1,65 +0,0 @@
|
||||
# -*- coding: utf-8 -*-
|
||||
"""
|
||||
f2py c_library.pyf c_functions.c -c
|
||||
|
||||
See also http://www.scipy.org/Cookbook/CompilingExtensionsOnWindowsWithMinGW
|
||||
"""
|
||||
import os
|
||||
import sys
|
||||
|
||||
|
||||
def which(program):
|
||||
"""
|
||||
Return filepath to program if it exists
|
||||
|
||||
In order to test if a certain executable exists, it will search for the
|
||||
program name in the environment variables.
|
||||
If program is a full path to an executable, it will check it exists
|
||||
|
||||
Copied from:
|
||||
http://stackoverflow.com/questions/377017/test-if-executable-exists-in-python/
|
||||
It is supposed to mimic the UNIX command "which"
|
||||
"""
|
||||
|
||||
def is_exe(fpath):
|
||||
return os.path.exists(fpath) and os.access(fpath, os.X_OK)
|
||||
|
||||
fpath, unused_fname = os.path.split(program)
|
||||
if fpath:
|
||||
if is_exe(program):
|
||||
return program
|
||||
else:
|
||||
for path in os.environ["PATH"].split(os.pathsep):
|
||||
exe_file = os.path.join(path, program)
|
||||
if is_exe(exe_file):
|
||||
return exe_file
|
||||
|
||||
return None
|
||||
|
||||
|
||||
def f2py_call_str():
|
||||
'''Return which f2py callable is in the path regardless of platform'''
|
||||
|
||||
# define possible options:
|
||||
# on Arch Linux, python and f2py are the calls corresponding to python 3
|
||||
# and python2/f2py2 for python 2
|
||||
# other Linux versions might still use python/f2py for python 2
|
||||
|
||||
if os.path.basename(sys.executable).endswith('2'):
|
||||
options = ('f2py2', 'f2py2.6', 'f2py2.7',)
|
||||
else: # on Windows and other Linux using python/f2py
|
||||
options = ('f2py.exe', 'f2py.bat', 'f2py', 'f2py2.6', 'f2py2.7',
|
||||
'f2py.py',)
|
||||
for k in options:
|
||||
if which(k):
|
||||
# Found the f2py path, no need to look further
|
||||
f2py_call = k
|
||||
f2py_path = which(k)
|
||||
break
|
||||
|
||||
try:
|
||||
print('found f2py in:', f2py_path)
|
||||
return f2py_call
|
||||
except NameError:
|
||||
raise UserWarning('Couldn\'t locate f2py. '
|
||||
'Should be part of NumPy installation.')
|
@ -1,908 +0,0 @@
|
||||
# /usr/bin/env python
|
||||
'''
|
||||
Module FIG
|
||||
------------
|
||||
Module for manipulating windows/figures created using
|
||||
pylab or enthought.mayavi.mlab on the windows platform.
|
||||
|
||||
Figure manipulation involves
|
||||
maximization, minimization, hiding, closing, stacking or tiling.
|
||||
|
||||
It is assumed that the figures are uniquely numbered in the following way:
|
||||
Figure 1
|
||||
Figure 2
|
||||
....
|
||||
or
|
||||
TVTK scene 1
|
||||
TVTK scene 2
|
||||
TVTK scene 3
|
||||
...
|
||||
|
||||
Example
|
||||
-------
|
||||
>>> import pylab as p
|
||||
>>> import wafo.fig as fig
|
||||
>>> for ix in range(6):
|
||||
... f = p.figure(ix)
|
||||
>>> fig.stack('all')
|
||||
>>> fig.stack(1,2)
|
||||
>>> fig.hide(1)
|
||||
>>> fig.restore(1)
|
||||
>>> fig.tile()
|
||||
>>> fig.pile()
|
||||
>>> fig.maximize(4)
|
||||
>>> fig.close('all')
|
||||
'''
|
||||
|
||||
from __future__ import absolute_import, division, print_function
|
||||
# import win32api
|
||||
import win32gui
|
||||
import win32con
|
||||
import wx
|
||||
import numpy
|
||||
|
||||
from win32gui import (EnumWindows, MoveWindow, GetWindowRect, FindWindow,
|
||||
ShowWindow, BringWindowToTop)
|
||||
|
||||
__all__ = ['close', 'cycle', 'hide', 'keep', 'maximize', 'minimize', 'pile',
|
||||
'restore', 'stack', 'tile', 'find_all_figure_numbers', 'set_size']
|
||||
|
||||
# Figure format strings to recognize in window title
|
||||
FIGURE_TITLE_FORMATS = ('Figure', 'TVTK Scene', 'Chaco Plot Window: Figure')
|
||||
_SCREENSIZE = None
|
||||
|
||||
|
||||
class CycleDialog(wx.Dialog):
|
||||
|
||||
def _get_buttons(self):
|
||||
hbox = wx.BoxSizer(wx.HORIZONTAL)
|
||||
buttons = ['Forward', 'Back', 'Cancel']
|
||||
callbacks = [self.on_forward, self.on_backward, self.on_cancel]
|
||||
for button, callback in zip(buttons, callbacks):
|
||||
button = wx.Button(self, -1, button, size=(70, 30))
|
||||
self.Bind(wx.EVT_BUTTON, callback, button)
|
||||
hbox.Add(button, 1, wx.ALIGN_CENTER)
|
||||
return hbox
|
||||
|
||||
def _get_message(self):
|
||||
label = ('Press back or forward to display previous or next figure(s),'
|
||||
' respectively. Press cancel to quit.')
|
||||
message = wx.StaticText(self, label=label, size=(240, 25))
|
||||
return message
|
||||
|
||||
def __init__(self, parent, interval=None, title='Cycle dialog'):
|
||||
super(CycleDialog, self).__init__(parent, title=title, size=(260, 130))
|
||||
if isinstance(interval, (float, int)):
|
||||
self.interval_milli_sec = interval * 1000
|
||||
else:
|
||||
self.interval_milli_sec = 30
|
||||
|
||||
self.timer = wx.Timer(self)
|
||||
self.Bind(wx.EVT_TIMER, self.on_forward, self.timer)
|
||||
|
||||
vbox = wx.BoxSizer(wx.VERTICAL)
|
||||
vbox.Add(self._get_message(), 0, wx.ALIGN_CENTER | wx.TOP, 20)
|
||||
vbox.Add(self._get_buttons(), 1, wx.ALIGN_CENTER | wx.TOP | wx.BOTTOM, 10)
|
||||
self.SetSizer(vbox)
|
||||
|
||||
def ShowModal(self, *args, **kwargs):
|
||||
self.timer.Start(self.interval_milli_sec, oneShot=True)
|
||||
return super(CycleDialog, self).ShowModal(*args, **kwargs)
|
||||
|
||||
def on_forward(self, evt):
|
||||
self.EndModal(wx.ID_FORWARD)
|
||||
|
||||
def on_backward(self, evt):
|
||||
self.EndModal(wx.ID_BACKWARD)
|
||||
|
||||
def on_cancel(self, evt):
|
||||
self.EndModal(wx.ID_CANCEL)
|
||||
|
||||
|
||||
def _get_cycle_dialog(parent=None, interval=None):
|
||||
app = wx.GetApp()
|
||||
if not app:
|
||||
app = wx.App(redirect=False)
|
||||
frame = wx.Frame(None)
|
||||
app.SetTopWindow(frame)
|
||||
dlg = CycleDialog(parent, interval)
|
||||
return dlg
|
||||
|
||||
|
||||
def get_window_position_and_size(window_handle):
|
||||
pos = GetWindowRect(window_handle)
|
||||
return pos[0], pos[1], pos[2] - pos[0], pos[3] - pos[1]
|
||||
|
||||
|
||||
def get_screen_position_and_size(window_handles):
|
||||
"""Return screen position; X, Y and size; width, height.
|
||||
|
||||
Parameters
|
||||
----------
|
||||
window_handles: list of handles to open window figures
|
||||
(Note: only needed the first time)
|
||||
|
||||
Returns
|
||||
--------
|
||||
X : coordinate of the left side of the screen.
|
||||
Y : coordinate of the top of the screen.
|
||||
width : screen horizontal size
|
||||
height : screen vertical size
|
||||
|
||||
"""
|
||||
# pylint: disable=global-statement
|
||||
global _SCREENSIZE
|
||||
if _SCREENSIZE is None:
|
||||
window_handle = window_handles[0]
|
||||
pos = get_window_position_and_size(window_handle)
|
||||
_show_windows((window_handle,), win32con.SW_SHOWMAXIMIZED)
|
||||
_SCREENSIZE = get_window_position_and_size(window_handle)
|
||||
MoveWindow(window_handle, pos[0], pos[1], pos[2], pos[3], 1)
|
||||
return _SCREENSIZE
|
||||
|
||||
|
||||
def _get_screen_size(wnds):
|
||||
screen_width, screen_height = get_screen_position_and_size(wnds)[2:4]
|
||||
return screen_width, screen_height
|
||||
|
||||
|
||||
def _windowEnumerationHandler(handle, result_list):
|
||||
"""Pass to win32gui.EnumWindows() to generate list of window handle, window
|
||||
text tuples."""
|
||||
# pylint: disable=no-member
|
||||
if win32gui.IsWindowVisible(handle):
|
||||
result_list.append((handle, win32gui.GetWindowText(handle)))
|
||||
|
||||
|
||||
def _find_window_handles_and_titles(wantedTitle=None):
|
||||
"""Return list of window handle and window title tuples.
|
||||
|
||||
Parameter
|
||||
---------
|
||||
wantedTitle:
|
||||
|
||||
"""
|
||||
handles_n_titles = []
|
||||
EnumWindows(_windowEnumerationHandler, handles_n_titles)
|
||||
if wantedTitle is None:
|
||||
return handles_n_titles
|
||||
else:
|
||||
return [(handle, title)
|
||||
for handle, title in handles_n_titles
|
||||
if title.startswith(wantedTitle)]
|
||||
|
||||
|
||||
def find_figure_handles(*figure_numbers):
|
||||
"""Find figure handles from figure numbers."""
|
||||
wnd_handles = []
|
||||
for figure_number in _parse_figure_numbers(*figure_numbers):
|
||||
for format_ in FIGURE_TITLE_FORMATS:
|
||||
winTitle = format_ + ' %d' % figure_number
|
||||
handle = FindWindow(None, winTitle)
|
||||
if not handle == 0:
|
||||
wnd_handles.append(handle)
|
||||
return wnd_handles
|
||||
|
||||
|
||||
def find_all_figure_numbers():
|
||||
"""Return list of all figure numbers.
|
||||
|
||||
Example
|
||||
-------
|
||||
>>> import fig
|
||||
>>> import pylab as p
|
||||
>>> for ix in range(5):
|
||||
... f = p.figure(ix)
|
||||
... p.draw()
|
||||
|
||||
fig.find_all_figure_numbers()
|
||||
[0, 1, 2, 3, 4]
|
||||
|
||||
>>> fig.close()
|
||||
|
||||
"""
|
||||
figure_numbers = []
|
||||
for wantedTitle in FIGURE_TITLE_FORMATS:
|
||||
handles_n_titles = _find_window_handles_and_titles(wantedTitle)
|
||||
for _handle, title in handles_n_titles:
|
||||
try:
|
||||
number = int(title.split()[-1])
|
||||
figure_numbers.append(number)
|
||||
except (TypeError, ValueError):
|
||||
pass
|
||||
# pylint: disable=no-member
|
||||
return numpy.unique(figure_numbers).tolist()
|
||||
|
||||
|
||||
def _parse_figure_numbers(*args):
|
||||
figure_numbers = []
|
||||
for arg in args:
|
||||
if isinstance(arg, (list, tuple, set)):
|
||||
for val in arg:
|
||||
figure_numbers.append(int(val))
|
||||
elif isinstance(arg, int):
|
||||
figure_numbers.append(arg)
|
||||
elif arg == 'all':
|
||||
figure_numbers = find_all_figure_numbers()
|
||||
break
|
||||
else:
|
||||
raise TypeError('Only integers arguments accepted!')
|
||||
|
||||
if len(figure_numbers) == 0:
|
||||
figure_numbers = find_all_figure_numbers()
|
||||
return figure_numbers
|
||||
|
||||
|
||||
def _show_figure(figure_numbers, command):
|
||||
"""Sets the specified figure's show state.
|
||||
|
||||
Parameters
|
||||
----------
|
||||
figure_numbers: list of figure numbers
|
||||
command: one of following commands:
|
||||
SW_FORCEMINIMIZE:
|
||||
Minimizes a window, even if the thread that owns the window is not
|
||||
responding. This flag should only be used when minimizing windows
|
||||
from a different thread.
|
||||
SW_HIDE:
|
||||
Hides the window and activates another window.
|
||||
SW_MAXIMIZE:
|
||||
Maximizes the specified window.
|
||||
SW_MINIMIZE:
|
||||
Minimizes the specified window and activates the next top-level window
|
||||
in the Z order.
|
||||
SW_RESTORE:
|
||||
Activates and displays the window. If the window is minimized or
|
||||
maximized, the system restores it to its original size and position.
|
||||
An application should specify this flag when restoring a minimized
|
||||
window.
|
||||
SW_SHOW:
|
||||
Activates the window and displays it in its current size and position.
|
||||
SW_SHOWDEFAULT:
|
||||
Sets the show state based on the SW_ value specified in the STARTUPINFO
|
||||
structure passed to the CreateProcess function by the program that
|
||||
started the application.
|
||||
SW_SHOWMAXIMIZED:
|
||||
Activates the window and displays it as a maximized window.
|
||||
SW_SHOWMINIMIZED:
|
||||
Activates the window and displays it as a minimized window.
|
||||
SW_SHOWMINNOACTIVE:
|
||||
Displays the window as a minimized window. This value is similar to
|
||||
SW_SHOWMINIMIZED, except the window is not activated.
|
||||
SW_SHOWNA:
|
||||
Displays the window in its current size and position. This value is
|
||||
similar to SW_SHOW, except the window is not activated.
|
||||
SW_SHOWNOACTIVATE:
|
||||
Displays a window in its most recent size and position. This value is
|
||||
similar to SW_SHOWNORMAL, except the window is not actived.
|
||||
SW_SHOWNORMAL:
|
||||
Activates and displays a window. If the window is minimized or
|
||||
maximized, the system restores it to its original size and position.
|
||||
An application should specify this flag when displaying the window for
|
||||
the first time.
|
||||
|
||||
"""
|
||||
for number in _parse_figure_numbers(*figure_numbers):
|
||||
for format_ in FIGURE_TITLE_FORMATS:
|
||||
title = format_ + ' %d' % number
|
||||
handle = FindWindow(None, title)
|
||||
if not handle == 0:
|
||||
BringWindowToTop(handle)
|
||||
ShowWindow(handle, command)
|
||||
|
||||
|
||||
def _show_windows(handles, command, redraw_now=False):
|
||||
"""Sets the specified window's show state.
|
||||
|
||||
Parameters
|
||||
----------
|
||||
handles: list of window handles
|
||||
command: one of following commands:
|
||||
SW_FORCEMINIMIZE:
|
||||
Minimizes a window, even if the thread that owns the window is not
|
||||
responding. This flag should only be used when minimizing windows
|
||||
from a different thread.
|
||||
SW_HIDE:
|
||||
Hides the window and activates another window.
|
||||
SW_MAXIMIZE:
|
||||
Maximizes the specified window.
|
||||
SW_MINIMIZE:
|
||||
Minimizes the specified window and activates the next top-level window
|
||||
in the Z order.
|
||||
SW_RESTORE:
|
||||
Activates and displays the window. If the window is minimized or
|
||||
maximized, the system restores it to its original size and position.
|
||||
An application should specify this flag when restoring a minimized
|
||||
window.
|
||||
SW_SHOW:
|
||||
Activates the window and displays it in its current size and position.
|
||||
SW_SHOWDEFAULT:
|
||||
Sets the show state based on the SW_ value specified in the STARTUPINFO
|
||||
structure passed to the CreateProcess function by the program that
|
||||
started the application.
|
||||
SW_SHOWMAXIMIZED:
|
||||
Activates the window and displays it as a maximized window.
|
||||
SW_SHOWMINIMIZED:
|
||||
Activates the window and displays it as a minimized window.
|
||||
SW_SHOWMINNOACTIVE:
|
||||
Displays the window as a minimized window. This value is similar to
|
||||
SW_SHOWMINIMIZED, except the window is not activated.
|
||||
SW_SHOWNA:
|
||||
Displays the window in its current size and position. This value is
|
||||
similar to SW_SHOW, except the window is not activated.
|
||||
SW_SHOWNOACTIVATE:
|
||||
Displays a window in its most recent size and position. This value is
|
||||
similar to SW_SHOWNORMAL, except the window is not actived.
|
||||
SW_SHOWNORMAL:
|
||||
Activates and displays a window. If the window is minimized or
|
||||
maximized, the system restores it to its original size and position.
|
||||
An application should specify this flag when displaying the window for
|
||||
the first time.
|
||||
|
||||
redraw_now :
|
||||
|
||||
"""
|
||||
# pylint: disable=no-member
|
||||
for handle in handles:
|
||||
if not handle == 0:
|
||||
BringWindowToTop(handle)
|
||||
ShowWindow(handle, command)
|
||||
if redraw_now:
|
||||
rect = GetWindowRect(handle)
|
||||
win32gui.RedrawWindow(handle, rect, None, win32con.RDW_UPDATENOW)
|
||||
|
||||
|
||||
def keep(*figure_numbers):
|
||||
"""Keeps figure windows of your choice and closes the rest.
|
||||
|
||||
Parameters
|
||||
----------
|
||||
figure_numbers : list of integers specifying which figures to keep.
|
||||
|
||||
Example:
|
||||
--------
|
||||
# keep only figures 1,2,3,5 and 7
|
||||
>>> import pylab as p
|
||||
>>> import wafo.fig as fig
|
||||
>>> for ix in range(10):
|
||||
... f = p.figure(ix)
|
||||
>>> fig.keep( range(1,4), 5, 7)
|
||||
|
||||
or
|
||||
fig.keep([range(1,4), 5, 7])
|
||||
>>> fig.close()
|
||||
|
||||
See also
|
||||
--------
|
||||
fig.close
|
||||
|
||||
"""
|
||||
figs2keep = []
|
||||
for fig in figure_numbers:
|
||||
if isinstance(fig, (list, tuple, set)):
|
||||
for val in fig:
|
||||
figs2keep.append(int(val))
|
||||
elif isinstance(fig, int):
|
||||
figs2keep.append(fig)
|
||||
else:
|
||||
raise TypeError('Only integers arguments accepted!')
|
||||
|
||||
if len(figs2keep) > 0:
|
||||
allfigs = set(find_all_figure_numbers())
|
||||
figs2delete = allfigs.difference(figs2keep)
|
||||
close(figs2delete)
|
||||
|
||||
|
||||
def close(*figure_numbers):
|
||||
""" Close figure window(s)
|
||||
|
||||
Parameters
|
||||
----------
|
||||
figure_numbers : list of integers or string
|
||||
specifying which figures to close (default 'all').
|
||||
|
||||
Examples
|
||||
--------
|
||||
>>> import pylab as p
|
||||
>>> import wafo.fig as fig
|
||||
>>> for ix in range(5):
|
||||
... f = p.figure(ix)
|
||||
>>> fig.close(3,4) # close figure 3 and 4
|
||||
>>> fig.close('all') # close all remaining figures
|
||||
|
||||
or even simpler
|
||||
fig.close() # close all remaining figures
|
||||
|
||||
See also
|
||||
--------
|
||||
fig.keep
|
||||
|
||||
"""
|
||||
# pylint: disable=no-member
|
||||
for handle in find_figure_handles(*figure_numbers):
|
||||
if win32gui.SendMessage(handle, win32con.WM_CLOSE, 0, 0):
|
||||
win32gui.SendMessage(handle, win32con.WM_DESTROY, 0, 0)
|
||||
|
||||
|
||||
def restore(*figure_numbers):
|
||||
"""Restore figures window size and position to its default value.
|
||||
|
||||
Parameters
|
||||
----------
|
||||
figure_numbers : list of integers or string
|
||||
specifying which figures to restor (default 'all').
|
||||
|
||||
Description
|
||||
-----------
|
||||
RESTORE Activates and displays the window. If the window is minimized
|
||||
or maximized, the system restores it to its original size and position.
|
||||
|
||||
Examples
|
||||
---------
|
||||
>>> import pylab as p
|
||||
>>> import wafo.fig as fig
|
||||
>>> for ix in range(5):
|
||||
... f = p.figure(ix)
|
||||
>>> fig.restore('all') #Restores all figures
|
||||
>>> fig.restore() #same as restore('all')
|
||||
>>> fig.restore(p.gcf().number) #Restores the current figure
|
||||
>>> fig.restore(3) #Restores figure 3
|
||||
>>> fig.restore([2, 4]) #Restores figures 2 and 4
|
||||
|
||||
or alternatively
|
||||
fig.restore(2, 4)
|
||||
>>> fig.close()
|
||||
|
||||
See also
|
||||
--------
|
||||
fig.close,
|
||||
fig.keep
|
||||
|
||||
"""
|
||||
SW_RESTORE = win32con.SW_RESTORE
|
||||
# SW_RESTORE = win32con.SW_SHOWDEFAULT
|
||||
# SW_RESTORE = win32con.SW_SHOWNORMAL
|
||||
_show_figure(figure_numbers, SW_RESTORE)
|
||||
|
||||
|
||||
def hide(*figure_numbers):
|
||||
"""hide figure(s) window.
|
||||
|
||||
Parameters
|
||||
----------
|
||||
figure_numbers : list of integers or string
|
||||
specifying which figures to hide (default 'all').
|
||||
|
||||
Examples:
|
||||
--------
|
||||
>>> import wafo.fig as fig
|
||||
>>> import pylab as p
|
||||
>>> for ix in range(5):
|
||||
... f = p.figure(ix)
|
||||
>>> fig.hide('all') #hides all unhidden figures
|
||||
>>> fig.hide() #same as hide('all')
|
||||
>>> fig.hide(p.gcf().number) #hides the current figure
|
||||
>>> fig.hide(3) #hides figure 3
|
||||
>>> fig.hide([2, 4]) #hides figures 2 and 4
|
||||
|
||||
or alternatively
|
||||
fig.hide(2, 4)
|
||||
>>> fig.restore(list(range(5)))
|
||||
>>> fig.close()
|
||||
|
||||
See also
|
||||
--------
|
||||
fig.cycle,
|
||||
fig.keep,
|
||||
fig.restore
|
||||
|
||||
"""
|
||||
_show_figure(figure_numbers, win32con.SW_HIDE)
|
||||
|
||||
|
||||
def minimize(*figure_numbers):
|
||||
"""Minimize figure(s) window size.
|
||||
|
||||
Parameters
|
||||
----------
|
||||
figure_numbers : list of integers or string
|
||||
specifying which figures to minimize (default 'all').
|
||||
|
||||
Examples:
|
||||
---------
|
||||
>>> import wafo.fig as fig
|
||||
>>> import pylab as p
|
||||
>>> for ix in range(5):
|
||||
... f = p.figure(ix)
|
||||
>>> fig.minimize('all') #Minimizes all unhidden figures
|
||||
>>> fig.minimize() #same as minimize('all')
|
||||
>>> fig.minimize(p.gcf().number) #Minimizes the current figure
|
||||
>>> fig.minimize(3) #Minimizes figure 3
|
||||
>>> fig.minimize([2, 4]) #Minimizes figures 2 and 4
|
||||
|
||||
or alternatively
|
||||
fig.minimize(2, 4)
|
||||
>>> fig.close()
|
||||
|
||||
See also
|
||||
--------
|
||||
fig.cycle,
|
||||
fig.keep,
|
||||
fig.restore
|
||||
|
||||
"""
|
||||
_show_figure(figure_numbers, win32con.SW_SHOWMINIMIZED)
|
||||
|
||||
|
||||
def maximize(*figure_numbers):
|
||||
"""Maximize figure(s) window size.
|
||||
|
||||
Parameters
|
||||
----------
|
||||
figure_numbers : list of integers or string
|
||||
specifying which figures to maximize (default 'all').
|
||||
|
||||
Examples:
|
||||
---------
|
||||
>>> import pylab as p
|
||||
>>> import wafo.fig as fig
|
||||
>>> for ix in range(5):
|
||||
... f = p.figure(ix)
|
||||
>>> fig.maximize('all') #Maximizes all unhidden figures
|
||||
>>> fig.maximize() #same as maximize('all')
|
||||
>>> fig.maximize(p.gcf().number) #Maximizes the current figure
|
||||
>>> fig.maximize(3) #Maximizes figure 3
|
||||
>>> fig.maximize([2, 4]) #Maximizes figures 2 and 4
|
||||
|
||||
or alternatively
|
||||
fig.maximize(2, 4)
|
||||
>>> fig.close()
|
||||
|
||||
See also
|
||||
--------
|
||||
fig.cycle,
|
||||
fig.keep,
|
||||
fig.restore
|
||||
|
||||
"""
|
||||
_show_figure(figure_numbers, win32con.SW_SHOWMAXIMIZED)
|
||||
|
||||
|
||||
def pile(*figure_numbers, **kwds):
|
||||
"""Pile figure windows.
|
||||
|
||||
Parameters
|
||||
----------
|
||||
figure_numbers : list of integers or string
|
||||
specifying which figures to pile (default 'all').
|
||||
kwds : dict with the following keys
|
||||
position :
|
||||
width :
|
||||
height :
|
||||
|
||||
Description
|
||||
-------------
|
||||
PILE piles all open figure windows on top of eachother
|
||||
with complete overlap. PILE(FIGS) can be used to specify which
|
||||
figures that should be piled. Figures are not sorted when specified.
|
||||
|
||||
Example:
|
||||
--------
|
||||
>>> import pylab as p
|
||||
>>> import wafo.fig as fig
|
||||
>>> for ix in range(7):
|
||||
... f = p.figure(ix)
|
||||
>>> fig.pile() # pile all open figures
|
||||
>>> fig.pile(range(1,4), 5, 7) # pile figure 1,2,3,5 and 7
|
||||
>>> fig.close()
|
||||
|
||||
See also
|
||||
--------
|
||||
fig.cycle, fig.keep, fig.maximize, fig.restore,
|
||||
fig.stack, fig.tile
|
||||
|
||||
"""
|
||||
wnds = find_figure_handles(*figure_numbers)
|
||||
numfigs = len(wnds)
|
||||
if numfigs > 0:
|
||||
screen_width, screen_height = _get_screen_size(wnds)
|
||||
pos = kwds.get(
|
||||
'position', (int(screen_width / 5), int(screen_height / 4)))
|
||||
width = kwds.get('width', int(screen_width / 2.5))
|
||||
height = kwds.get('height', int(screen_height / 2))
|
||||
|
||||
for wnd in wnds:
|
||||
MoveWindow(wnd, pos[0], pos[1], width, height, 1)
|
||||
BringWindowToTop(wnd)
|
||||
|
||||
|
||||
def set_size(*figure_numbers, **kwds):
|
||||
"""Set size for figure windows.
|
||||
|
||||
Parameters
|
||||
----------
|
||||
figure_numbers : list of integers or string
|
||||
specifying which figures to pile (default 'all').
|
||||
kwds : dict with the following keys
|
||||
width :
|
||||
height :
|
||||
|
||||
Description
|
||||
-------------
|
||||
Set size sets the size of all open figure windows. SET_SIZE(FIGS)
|
||||
can be used to specify which figures that should be resized.
|
||||
Figures are not sorted when specified.
|
||||
|
||||
Example:
|
||||
--------
|
||||
>>> import pylab as p
|
||||
>>> import fig
|
||||
>>> for ix in range(7):
|
||||
... f = p.figure(ix)
|
||||
>>> fig.set_size(7, width=150, height=100)
|
||||
>>> fig.set_size(range(1,4), 5,width=250, height=170)
|
||||
>>> fig.close()
|
||||
|
||||
See also
|
||||
--------
|
||||
fig.cycle, fig.keep, fig.maximize, fig.restore,
|
||||
fig.stack, fig.tile
|
||||
|
||||
"""
|
||||
handles = find_figure_handles(*figure_numbers)
|
||||
numfigs = len(handles)
|
||||
if numfigs > 0:
|
||||
screen_width, screen_height = _get_screen_size(handles)
|
||||
width = kwds.get('width', int(screen_width / 2.5))
|
||||
height = kwds.get('height', int(screen_height / 2))
|
||||
new_pos = kwds.get('position', None)
|
||||
pos = new_pos
|
||||
for handle in handles:
|
||||
if not new_pos:
|
||||
pos = get_window_position_and_size(handle)
|
||||
MoveWindow(handle, pos[0], pos[1], width, height, 1)
|
||||
BringWindowToTop(handle)
|
||||
|
||||
|
||||
def stack(*figure_numbers, **kwds):
|
||||
"""Stack figure windows.
|
||||
|
||||
Parameters
|
||||
----------
|
||||
figure_numbers : list of integers or string
|
||||
specifying which figures to stack (default 'all').
|
||||
kwds : dict with the following keys
|
||||
figs_per_stack :
|
||||
number of figures per stack (default depends on screenheight)
|
||||
|
||||
Description
|
||||
-----------
|
||||
STACK stacks all open figure windows on top of eachother
|
||||
with maximum overlap. STACK(FIGS) can be used to specify which
|
||||
figures that should be stacked. Figures are not sorted when specified.
|
||||
|
||||
Example:
|
||||
--------
|
||||
>>> import pylab as p
|
||||
>>> import wafo.fig as fig
|
||||
>>> for ix in range(7):
|
||||
... f = p.figure(ix)
|
||||
>>> fig.stack() # stack all open figures
|
||||
>>> fig.stack(range(1,4), 5, 7) # stack figure 1,2,3,5 and 7
|
||||
>>> fig.close()
|
||||
|
||||
See also
|
||||
--------
|
||||
fig.cycle, fig.keep, fig.maximize, fig.restore,
|
||||
fig.pile, fig.tile
|
||||
|
||||
"""
|
||||
wnds = find_figure_handles(*figure_numbers)
|
||||
numfigs = len(wnds)
|
||||
if numfigs > 0:
|
||||
screenpos = get_screen_position_and_size(wnds)
|
||||
y_step = 25
|
||||
x_step = border = 5
|
||||
|
||||
figs_per_stack = kwds.get(
|
||||
'figs_per_stack',
|
||||
int(numpy.fix(0.7 * (screenpos[3] - border) / y_step)))
|
||||
|
||||
for iy in range(numfigs):
|
||||
pos = get_window_position_and_size(wnds[iy])
|
||||
# print('[x, y, w, h] = ', pos)
|
||||
ix = iy % figs_per_stack
|
||||
ypos = int(screenpos[1] + ix * y_step + border)
|
||||
xpos = int(screenpos[0] + ix * x_step + border)
|
||||
MoveWindow(wnds[iy], xpos, ypos, pos[2], pos[3], 1)
|
||||
BringWindowToTop(wnds[iy])
|
||||
|
||||
|
||||
def tile(*figure_numbers, **kwds):
|
||||
"""Tile figure windows.
|
||||
|
||||
Parameters
|
||||
----------
|
||||
figure_numbers : list of integers or string
|
||||
specifying which figures to tile (default 'all').
|
||||
kwds : dict with key pairs
|
||||
specifying how many pairs of figures that are tiled at a time
|
||||
|
||||
Description
|
||||
-----------
|
||||
TILE places all open figure windows around on the screen with no
|
||||
overlap. TILE(FIGS) can be used to specify which figures that
|
||||
should be tiled. Figures are not sorted when specified.
|
||||
|
||||
Example:
|
||||
--------
|
||||
>>> import pylab as p
|
||||
>>> import wafo.fig as fig
|
||||
>>> for ix in range(7):
|
||||
... f = p.figure(ix)
|
||||
>>> fig.tile() # tile all open figures
|
||||
>>> fig.tile( range(1,4), 5, 7) # tile figure 1,2,3,5 and 7
|
||||
>>> fig.tile(range(1,11), pairs=2) # tile figure 1 to 10 two at a time
|
||||
>>> fig.tile(range(1,11), pairs=3) # tile figure 1 to 10 three at a time
|
||||
>>> fig.close()
|
||||
|
||||
See also
|
||||
--------
|
||||
fig.cycle, fig.keep, fig.maximize, fig.minimize
|
||||
fig.restore, fig.pile, fig.stack
|
||||
|
||||
"""
|
||||
wnds = find_figure_handles(*figure_numbers)
|
||||
|
||||
nfigs = len(wnds)
|
||||
# Number of windows.
|
||||
|
||||
if nfigs > 0:
|
||||
nfigspertile = kwds.get('pairs', nfigs)
|
||||
|
||||
ceil = numpy.ceil
|
||||
sqrt = numpy.sqrt
|
||||
maximum = numpy.maximum
|
||||
|
||||
nlayers = int(ceil(nfigs / nfigspertile))
|
||||
|
||||
# Number of figures horisontally.
|
||||
nh = maximum(int(ceil(sqrt(nfigspertile))), 2)
|
||||
# Number of figures vertically.
|
||||
nv = maximum(int(ceil(nfigspertile / nh)), 2)
|
||||
|
||||
screenpos = get_screen_position_and_size(wnds)
|
||||
screen_width, screen_heigth = screenpos[2:4]
|
||||
|
||||
hspc = 10 # Horisontal space.
|
||||
topspc = 20 # Space above top figure.
|
||||
medspc = 10 # Space between figures.
|
||||
botspc = 20 # Space below bottom figure.
|
||||
|
||||
figwid = (screen_width - (nh + 1) * hspc) / nh
|
||||
fighgt = (screen_heigth - (topspc + botspc) - (nv - 1) * medspc) / nv
|
||||
|
||||
figwid = int(numpy.round(figwid))
|
||||
fighgt = int(numpy.round(fighgt))
|
||||
|
||||
idx = 0
|
||||
for unused_ix in range(nlayers):
|
||||
for row in range(nv):
|
||||
figtop = int(screenpos[1] + topspc + row * (fighgt + medspc))
|
||||
for col in range(nh):
|
||||
if (row) * nh + col < nfigspertile:
|
||||
if idx < nfigs:
|
||||
figlft = int(
|
||||
screenpos[0] + (col + 1) * hspc + col * figwid)
|
||||
fighnd = wnds[idx]
|
||||
MoveWindow(fighnd, figlft, figtop, figwid, fighgt,
|
||||
1)
|
||||
# Set position.
|
||||
BringWindowToTop(fighnd)
|
||||
idx += 1
|
||||
|
||||
|
||||
class _CycleGenerator(object):
|
||||
|
||||
"""Cycle through figure windows.
|
||||
|
||||
Parameters
|
||||
----------
|
||||
figure_numbers : list of integers or string
|
||||
specifying which figures to cycle through (default 'all').
|
||||
kwds : dict with the following keys
|
||||
pairs : number of figures to cycle in pairs (default 1)
|
||||
maximize: If True maximize figure when viewing (default False)
|
||||
interval : pause interval in seconds
|
||||
|
||||
Description
|
||||
-----------
|
||||
CYCLE brings up all open figure in ascending order and pauses after
|
||||
each figure. Press escape to quit cycling, backspace to display previous
|
||||
figure(s) and press any other key to display next figure(s)
|
||||
When done, the figures are sorted in ascending order.
|
||||
|
||||
CYCLE(maximize=True) does the same thing, except figures are maximized.
|
||||
CYCLE(pairs=2) cycle through all figures in pairs of 2.
|
||||
|
||||
Examples:
|
||||
>>> import pylab as p
|
||||
>>> import wafo.fig as fig
|
||||
>>> for ix in range(4):
|
||||
... f = p.figure(ix)
|
||||
|
||||
fig.cycle(range(3), interval=1) # Cycle trough figure 0 to 2
|
||||
|
||||
# Cycle trough figure 0 to 2 with figures maximized
|
||||
fig.cycle(range(3), maximize=True, interval=1)
|
||||
fig.cycle(interval=1) # Cycle through all figures one at a time
|
||||
fig.tile(pairs=2, interval=1)
|
||||
fig.cycle(pairs=2, interval=2) # Cycle through all figures two at a time
|
||||
|
||||
fig.cycle(pairs=2) # Manually cycle through all figures two at a time
|
||||
>>> fig.close()
|
||||
|
||||
See also
|
||||
--------
|
||||
fig.keep, fig.maximize, fig.restore, fig.pile,
|
||||
fig.stack, fig.tile
|
||||
|
||||
"""
|
||||
escape_key = chr(27)
|
||||
backspace_key = chr(8)
|
||||
|
||||
def __init__(self, **kwds):
|
||||
self.dialog = None
|
||||
maximize = kwds.get('maximize', False)
|
||||
pairs = kwds.get('pairs', 1)
|
||||
self.interval = kwds.get('interval', 'user_defined')
|
||||
self.nfigspercycle = 1
|
||||
if maximize:
|
||||
self.command = win32con.SW_SHOWMAXIMIZED
|
||||
else:
|
||||
self.command = win32con.SW_SHOWNORMAL
|
||||
if pairs is not None:
|
||||
self.nfigspercycle = pairs
|
||||
|
||||
def _set_options(self, kwds):
|
||||
self.__init__(**kwds)
|
||||
|
||||
def _iterate(self, handles):
|
||||
i = 0
|
||||
numfigs = len(handles)
|
||||
self.dialog = _get_cycle_dialog(parent=None, interval=self.interval)
|
||||
while 0 <= i and i < numfigs:
|
||||
iu = min(i + self.nfigspercycle, numfigs)
|
||||
yield handles[i:iu]
|
||||
i = self.next_index(i)
|
||||
self.dialog.Destroy()
|
||||
raise StopIteration
|
||||
|
||||
def next_index(self, i):
|
||||
result = self.dialog.ShowModal()
|
||||
if result == wx.ID_FORWARD:
|
||||
i += self.nfigspercycle
|
||||
elif result == wx.ID_BACKWARD:
|
||||
i -= self.nfigspercycle
|
||||
else:
|
||||
i = -1
|
||||
return i
|
||||
|
||||
def __call__(self, *figure_numbers, **kwds):
|
||||
handles = find_figure_handles(*figure_numbers)
|
||||
numfigs = len(handles)
|
||||
if numfigs > 0:
|
||||
self._set_options(kwds)
|
||||
for handle in self._iterate(handles):
|
||||
_show_windows(handle, self.command, redraw_now=True)
|
||||
|
||||
_show_windows(handles, win32con.SW_SHOWNORMAL)
|
||||
|
||||
cycle = _CycleGenerator()
|
||||
|
||||
|
||||
if __name__ == '__main__':
|
||||
from wafo.testing import test_docstrings
|
||||
import matplotlib
|
||||
matplotlib.interactive(True)
|
||||
test_docstrings(__file__)
|
File diff suppressed because it is too large
Load Diff
File diff suppressed because it is too large
Load Diff
@ -1,538 +0,0 @@
|
||||
"""
|
||||
Created on 20. aug. 2015
|
||||
|
||||
@author: pab
|
||||
"""
|
||||
from __future__ import division
|
||||
from collections import namedtuple
|
||||
import warnings
|
||||
import numdifftools as nd
|
||||
import numdifftools.nd_algopy as nda
|
||||
from numdifftools.extrapolation import dea3
|
||||
from numdifftools.limits import Limit
|
||||
import numpy as np
|
||||
from numpy import linalg
|
||||
from numpy.polynomial.chebyshev import chebval, Chebyshev
|
||||
from numpy.polynomial import polynomial
|
||||
from wafo.misc import piecewise, findcross, ecross
|
||||
|
||||
_FINFO = np.finfo(float)
|
||||
EPS = _FINFO.eps
|
||||
_EPS = EPS
|
||||
_TINY = _FINFO.tiny
|
||||
_HUGE = _FINFO.max
|
||||
|
||||
|
||||
def _assert(cond, msg):
|
||||
if not cond:
|
||||
raise ValueError(msg)
|
||||
|
||||
|
||||
def _assert_warn(cond, msg):
|
||||
if not cond:
|
||||
warnings.warn(msg)
|
||||
|
||||
|
||||
class PolyBasis(object):
|
||||
@staticmethod
|
||||
def _derivative(c, m):
|
||||
return polynomial.polyder(c, m)
|
||||
|
||||
@staticmethod
|
||||
def eval(t, c):
|
||||
return polynomial.polyval(t, c)
|
||||
|
||||
@staticmethod
|
||||
def _coefficients(k):
|
||||
c = np.zeros(k + 1)
|
||||
c[k] = 1
|
||||
return c
|
||||
|
||||
def derivative(self, t, k, n=1):
|
||||
c = self._coefficients(k)
|
||||
d_c = self._derivative(c, m=n)
|
||||
return self.eval(t, d_c)
|
||||
|
||||
def __call__(self, t, k):
|
||||
return t**k
|
||||
|
||||
|
||||
poly_basis = PolyBasis()
|
||||
|
||||
|
||||
class ChebyshevBasis(PolyBasis):
|
||||
|
||||
@staticmethod
|
||||
def _derivative(c, m):
|
||||
cheb = Chebyshev(c)
|
||||
dcheb = cheb.deriv(m=m)
|
||||
return dcheb.coef
|
||||
|
||||
@staticmethod
|
||||
def eval(t, c):
|
||||
return chebval(t, c)
|
||||
|
||||
def __call__(self, t, k):
|
||||
c = self._coefficients(k)
|
||||
return self.eval(t, c)
|
||||
|
||||
|
||||
chebyshev_basis = ChebyshevBasis()
|
||||
|
||||
|
||||
def richardson(q_val, k):
|
||||
# license BSD
|
||||
# Richardson extrapolation with parameter estimation
|
||||
c = np.real((q_val[k - 1] - q_val[k - 2]) / (q_val[k] - q_val[k - 1])) - 1.
|
||||
# The lower bound 0.07 admits the singularity x.^-0.9
|
||||
c = max(c, 0.07)
|
||||
return q_val[k] + (q_val[k] - q_val[k - 1]) / c
|
||||
|
||||
|
||||
def evans_webster_weights(omega, g, d_g, x, basis, *args, **kwds):
|
||||
|
||||
def psi(t, k):
|
||||
return d_g(t, *args, **kwds) * basis(t, k)
|
||||
|
||||
j_w = 1j * omega
|
||||
n = len(x)
|
||||
a_matrix = np.zeros((n, n), dtype=complex)
|
||||
rhs = np.zeros((n,), dtype=complex)
|
||||
|
||||
dbasis = basis.derivative
|
||||
lim_g = Limit(g)
|
||||
b_1 = np.exp(j_w * lim_g(1, *args, **kwds))
|
||||
if np.isnan(b_1):
|
||||
b_1 = 0.0
|
||||
a_1 = np.exp(j_w * lim_g(-1, *args, **kwds))
|
||||
if np.isnan(a_1):
|
||||
a_1 = 0.0
|
||||
|
||||
lim_psi = Limit(psi)
|
||||
for k in range(n):
|
||||
rhs[k] = basis(1, k) * b_1 - basis(-1, k) * a_1
|
||||
a_matrix[k] = (dbasis(x, k, n=1) + j_w * lim_psi(x, k))
|
||||
|
||||
solution = linalg.lstsq(a_matrix, rhs)
|
||||
return solution[0]
|
||||
|
||||
|
||||
def osc_weights(omega, g, d_g, x, basis, a_b, *args, **kwds):
|
||||
def _g(t):
|
||||
return g(scale * t + offset, *args, **kwds)
|
||||
|
||||
def _d_g(t):
|
||||
return scale * d_g(scale * t + offset, *args, **kwds)
|
||||
|
||||
w = []
|
||||
|
||||
for a, b in zip(a_b[::2], a_b[1::2]):
|
||||
scale = (b - a) / 2
|
||||
offset = (a + b) / 2
|
||||
|
||||
w.append(evans_webster_weights(omega, _g, _d_g, x, basis))
|
||||
|
||||
return np.asarray(w).ravel()
|
||||
|
||||
|
||||
class _Integrator(object):
|
||||
info = namedtuple('info', ['error_estimate', 'n'])
|
||||
|
||||
def __init__(self, f, g, dg=None, a=-1, b=1, basis=chebyshev_basis, s=1,
|
||||
precision=10, endpoints=True, full_output=False):
|
||||
self.f = f
|
||||
self.g = g
|
||||
self.dg = nd.Derivative(g) if dg is None else dg
|
||||
self.basis = basis
|
||||
self.a = a
|
||||
self.b = b
|
||||
self.s = s
|
||||
self.endpoints = endpoints
|
||||
self.precision = precision
|
||||
self.full_output = full_output
|
||||
|
||||
|
||||
class QuadOsc(_Integrator):
|
||||
def __init__(self, f, g, dg=None, a=-1, b=1, basis=chebyshev_basis, s=15,
|
||||
precision=10, endpoints=False, full_output=False, maxiter=17):
|
||||
self.maxiter = maxiter
|
||||
super(QuadOsc, self).__init__(f, g, dg=dg, a=a, b=b, basis=basis, s=s,
|
||||
precision=precision, endpoints=endpoints,
|
||||
full_output=full_output)
|
||||
|
||||
@staticmethod
|
||||
def _change_interval_to_0_1(f, g, d_g, a, _b):
|
||||
def f_01(t, *args, **kwds):
|
||||
den = 1 - t
|
||||
return f(a + t / den, *args, **kwds) / den ** 2
|
||||
|
||||
def g_01(t, *args, **kwds):
|
||||
return g(a + t / (1 - t), *args, **kwds)
|
||||
|
||||
def d_g_01(t, *args, **kwds):
|
||||
den = 1 - t
|
||||
return d_g(a + t / den, *args, **kwds) / den ** 2
|
||||
return f_01, g_01, d_g_01, 0., 1.
|
||||
|
||||
@staticmethod
|
||||
def _change_interval_to_m1_0(f, g, d_g, _a, b):
|
||||
def f_m10(t, *args, **kwds):
|
||||
den = 1 + t
|
||||
return f(b + t / den, *args, **kwds) / den ** 2
|
||||
|
||||
def g_m10(t, *args, **kwds):
|
||||
return g(b + t / (1 + t), *args, **kwds)
|
||||
|
||||
def d_g_m10(t, *args, **kwds):
|
||||
den = 1 + t
|
||||
return d_g(b + t / den, *args, **kwds) / den ** 2
|
||||
return f_m10, g_m10, d_g_m10, -1.0, 0.0
|
||||
|
||||
@staticmethod
|
||||
def _change_interval_to_m1_1(f, g, d_g, _a, _b):
|
||||
def f_m11(t, *args, **kwds):
|
||||
den = (1 - t**2)
|
||||
return f(t / den, *args, **kwds) * (1 + t**2) / den ** 2
|
||||
|
||||
def g_m11(t, *args, **kwds):
|
||||
den = (1 - t**2)
|
||||
return g(t / den, *args, **kwds)
|
||||
|
||||
def d_g_m11(t, *args, **kwds):
|
||||
den = (1 - t**2)
|
||||
return d_g(t / den, *args, **kwds) * (1 + t**2) / den ** 2
|
||||
return f_m11, g_m11, d_g_m11, -1., 1.
|
||||
|
||||
def _get_functions(self):
|
||||
a, b = self.a, self.b
|
||||
reverse = np.real(a) > np.real(b)
|
||||
if reverse:
|
||||
a, b = b, a
|
||||
f, g, dg = self.f, self.g, self.dg
|
||||
|
||||
if a == b:
|
||||
pass
|
||||
elif np.isinf(a) | np.isinf(b):
|
||||
# Check real limits
|
||||
if ~np.isreal(a) | ~np.isreal(b) | np.isnan(a) | np.isnan(b):
|
||||
raise ValueError('Infinite intervals must be real.')
|
||||
# Change of variable
|
||||
if np.isfinite(a) & np.isinf(b):
|
||||
f, g, dg, a, b = self._change_interval_to_0_1(f, g, dg, a, b)
|
||||
elif np.isinf(a) & np.isfinite(b):
|
||||
f, g, dg, a, b = self._change_interval_to_m1_0(f, g, dg, a, b)
|
||||
else: # -inf to inf
|
||||
f, g, dg, a, b = self._change_interval_to_m1_1(f, g, dg, a, b)
|
||||
|
||||
return f, g, dg, a, b, reverse
|
||||
|
||||
def __call__(self, omega, *args, **kwds):
|
||||
f, g, dg, a, b, reverse = self._get_functions()
|
||||
|
||||
val, err = self._quad_osc(f, g, dg, a, b, omega, *args, **kwds)
|
||||
|
||||
if reverse:
|
||||
val = -val
|
||||
if self.full_output:
|
||||
return val, err
|
||||
return val
|
||||
|
||||
@staticmethod
|
||||
def _get_best_estimate(k, q_0, q_1, q_2):
|
||||
if k >= 5:
|
||||
q_v = np.hstack((q_0[k], q_1[k], q_2[k]))
|
||||
q_w = np.hstack((q_0[k - 1], q_1[k - 1], q_2[k - 1]))
|
||||
elif k >= 3:
|
||||
q_v = np.hstack((q_0[k], q_1[k]))
|
||||
q_w = np.hstack((q_0[k - 1], q_1[k - 1]))
|
||||
else:
|
||||
q_v = np.atleast_1d(q_0[k])
|
||||
q_w = q_0[k - 1]
|
||||
errors = np.atleast_1d(abs(q_v - q_w))
|
||||
j = np.nanargmin(errors)
|
||||
return q_v[j], errors[j]
|
||||
|
||||
def _extrapolate(self, k, q_0, q_1, q_2):
|
||||
if k >= 4:
|
||||
q_1[k] = dea3(q_0[k - 2], q_0[k - 1], q_0[k])[0]
|
||||
q_2[k] = dea3(q_1[k - 2], q_1[k - 1], q_1[k])[0]
|
||||
elif k >= 2:
|
||||
q_1[k] = dea3(q_0[k - 2], q_0[k - 1], q_0[k])[0]
|
||||
# # Richardson extrapolation
|
||||
# if k >= 4:
|
||||
# q_1[k] = richardson(q_0, k)
|
||||
# q_2[k] = richardson(q_1, k)
|
||||
# elif k >= 2:
|
||||
# q_1[k] = richardson(q_0, k)
|
||||
q, err = self._get_best_estimate(k, q_0, q_1, q_2)
|
||||
return q, err
|
||||
|
||||
def _quad_osc(self, f, g, dg, a, b, omega, *args, **kwds):
|
||||
if a == b:
|
||||
q_val = b - a
|
||||
err = np.abs(b - a)
|
||||
return q_val, err
|
||||
|
||||
abseps = 10**-self.precision
|
||||
max_iter = self.maxiter
|
||||
basis = self.basis
|
||||
if self.endpoints:
|
||||
x_n = chebyshev_extrema(self.s)
|
||||
else:
|
||||
x_n = chebyshev_roots(self.s)
|
||||
# x_n = tanh_sinh_open_nodes(self.s)
|
||||
|
||||
# One interval
|
||||
hh = (b - a) / 2
|
||||
x = (a + b) / 2 + hh * x_n # Nodes
|
||||
|
||||
dtype = complex
|
||||
val0 = np.zeros((max_iter, 1), dtype=dtype) # Quadrature
|
||||
val1 = np.zeros((max_iter, 1), dtype=dtype) # First extrapolation
|
||||
val2 = np.zeros((max_iter, 1), dtype=dtype) # Second extrapolation
|
||||
|
||||
lim_f = Limit(f)
|
||||
a_b = np.hstack([a, b])
|
||||
wq = osc_weights(omega, g, dg, x_n, basis, a_b, *args, **kwds)
|
||||
val0[0] = hh * np.sum(wq * lim_f(x, *args, **kwds))
|
||||
|
||||
# Successive bisection of intervals
|
||||
nq = len(x_n)
|
||||
n = nq
|
||||
for k in range(1, max_iter):
|
||||
n += nq * 2**k
|
||||
|
||||
hh = hh / 2
|
||||
x = np.hstack([x + a, x + b]) / 2
|
||||
a_b = np.hstack([a_b + a, a_b + b]) / 2
|
||||
wq = osc_weights(omega, g, dg, x_n, basis, a_b, *args, **kwds)
|
||||
|
||||
val0[k] = hh * np.sum(wq * lim_f(x, *args, **kwds))
|
||||
|
||||
q_val, err = self._extrapolate(k, val0, val1, val2)
|
||||
|
||||
converged = (err <= abseps) | ~np.isfinite(q_val)
|
||||
if converged:
|
||||
break
|
||||
_assert_warn(converged, 'Max number of iterations reached '
|
||||
'without convergence.')
|
||||
_assert_warn(np.isfinite(q_val),
|
||||
'Integral approximation is Infinite or NaN.')
|
||||
|
||||
# The error estimate should not be zero
|
||||
err += 2 * np.finfo(q_val).eps
|
||||
return q_val, self.info(err, n)
|
||||
|
||||
|
||||
def adaptive_levin_points(m, delta):
|
||||
m_1 = m - 1
|
||||
prm = 0.5
|
||||
while prm * m_1 / delta >= 1:
|
||||
delta = 2 * delta
|
||||
k = np.arange(m)
|
||||
x = piecewise([k < prm * m_1, k == np.ceil(prm * m_1)],
|
||||
[-1 + k / delta, 0 * k, 1 - (m_1 - k) / delta])
|
||||
return x
|
||||
|
||||
|
||||
def open_levin_points(m, delta):
|
||||
return adaptive_levin_points(m + 2, delta)[1:-1]
|
||||
|
||||
|
||||
def chebyshev_extrema(m, delta=None):
|
||||
k = np.arange(m)
|
||||
x = np.cos(k * np.pi / (m - 1))
|
||||
return x
|
||||
|
||||
|
||||
def tanh_sinh_nodes(m, delta=None, tol=_EPS):
|
||||
tmax = np.arcsinh(np.arctanh(1 - _EPS) * 2 / np.pi)
|
||||
# tmax = 3.18
|
||||
m_1 = int(np.floor(-np.log2(tmax / max(m - 1, 1)))) - 1
|
||||
h = 2.0**-m_1
|
||||
t = np.arange((m + 1) // 2 + 1) * h
|
||||
x = np.tanh(np.pi / 2 * np.sinh(t))
|
||||
k = np.flatnonzero(np.abs(x - 1) <= 10 * tol)
|
||||
y = x[:k[0] + 1] if len(k) else x
|
||||
return np.hstack((-y[:0:-1], y))
|
||||
|
||||
|
||||
def tanh_sinh_open_nodes(m, delta=None, tol=_EPS):
|
||||
return tanh_sinh_nodes(m + 1, delta, tol)[1:-1]
|
||||
|
||||
|
||||
def chebyshev_roots(m, delta=None):
|
||||
k = np.arange(1, 2 * m, 2) * 0.5
|
||||
x = np.cos(k * np.pi / m)
|
||||
return x
|
||||
|
||||
|
||||
class AdaptiveLevin(_Integrator):
|
||||
"""Return integral for the Levin-type and adaptive Levin-type methods"""
|
||||
|
||||
@staticmethod
|
||||
def _a_levin(omega, f, g, d_g, x, s, basis, *args, **kwds):
|
||||
|
||||
def psi(t, k):
|
||||
return d_g(t, *args, **kwds) * basis(t, k)
|
||||
|
||||
j_w = 1j * omega
|
||||
nu = np.ones((len(x),), dtype=int)
|
||||
nu[0] = nu[-1] = s
|
||||
S = np.cumsum(np.hstack((nu, 0)))
|
||||
S[-1] = 0
|
||||
n = int(S[-2])
|
||||
a_matrix = np.zeros((n, n), dtype=complex)
|
||||
rhs = np.zeros((n,))
|
||||
dff = Limit(nda.Derivative(f))
|
||||
d_psi = Limit(nda.Derivative(psi))
|
||||
dbasis = basis.derivative
|
||||
for r, t in enumerate(x):
|
||||
for j in range(S[r - 1], S[r]):
|
||||
order = ((j - S[r - 1]) % nu[r]) # derivative order
|
||||
dff.fun.n = order
|
||||
rhs[j] = dff(t, *args, **kwds)
|
||||
d_psi.fun.n = order
|
||||
for k in range(n):
|
||||
a_matrix[j, k] = (dbasis(t, k, n=order + 1) +
|
||||
j_w * d_psi(t, k))
|
||||
k1 = np.flatnonzero(1 - np.isfinite(rhs))
|
||||
if k1.size > 0: # Remove singularities
|
||||
warnings.warn('Singularities detected! ')
|
||||
a_matrix[k1] = 0
|
||||
rhs[k1] = 0
|
||||
solution = linalg.lstsq(a_matrix, rhs)
|
||||
v = basis.eval([-1, 1], solution[0])
|
||||
|
||||
lim_g = Limit(g)
|
||||
g_b = np.exp(j_w * lim_g(1, *args, **kwds))
|
||||
if np.isnan(g_b):
|
||||
g_b = 0
|
||||
g_a = np.exp(j_w * lim_g(-1, *args, **kwds))
|
||||
if np.isnan(g_a):
|
||||
g_a = 0
|
||||
return v[1] * g_b - v[0] * g_a
|
||||
|
||||
def _get_integration_limits(self, omega, args, kwds):
|
||||
a, b = self.a, self.b
|
||||
M = 30
|
||||
ab = [a]
|
||||
scale = (b - a) / 2
|
||||
n = 30
|
||||
x = np.linspace(a, b, n + 1)
|
||||
dg_x = np.asarray([scale * omega * self.dg(xi, *args, **kwds)
|
||||
for xi in x])
|
||||
i10 = findcross(dg_x, M)
|
||||
i1 = findcross(dg_x, 1)
|
||||
i0 = findcross(dg_x, 0)
|
||||
im1 = findcross(dg_x, -1)
|
||||
im10 = findcross(dg_x, -M)
|
||||
x10 = ecross(x, dg_x, i10, M) if len(i10) else ()
|
||||
x1 = ecross(x, dg_x, i1, 1) if len(i1) else ()
|
||||
x0 = ecross(x, dg_x, i0, 0) if len(i0) else ()
|
||||
xm1 = ecross(x, dg_x, im1, -1) if len(im1) else ()
|
||||
xm10 = ecross(x, dg_x, im10, -M) if len(im10) else ()
|
||||
|
||||
for i in np.unique(np.hstack((x10, x1, x0, xm1, xm10))):
|
||||
if x[0] < i < x[n]:
|
||||
ab.append(i)
|
||||
ab.append(b)
|
||||
return ab
|
||||
|
||||
def __call__(self, omega, *args, **kwds):
|
||||
ab = self._get_integration_limits(omega, args, kwds)
|
||||
s = self.s
|
||||
val = 0
|
||||
n = 0
|
||||
err = 0
|
||||
for ai, bi in zip(ab[:-1], ab[1:]):
|
||||
vali, infoi = self._QaL(s, ai, bi, omega, *args, **kwds)
|
||||
val += vali
|
||||
err += infoi.error_estimate
|
||||
n += infoi.n
|
||||
if self.full_output:
|
||||
info = self.info(err, n)
|
||||
return val, info
|
||||
return val
|
||||
|
||||
@staticmethod
|
||||
def _get_num_points(s, prec, betam):
|
||||
return 1 if s > 1 else int(prec / max(np.log10(betam + 1), 1) + 1)
|
||||
|
||||
def _QaL(self, s, a, b, omega, *args, **kwds):
|
||||
"""if s>1,the integral is computed by Q_s^L"""
|
||||
scale = (b - a) / 2
|
||||
offset = (a + b) / 2
|
||||
prec = self.precision # desired precision
|
||||
|
||||
def ff(t, *args, **kwds):
|
||||
return scale * self.f(scale * t + offset, *args, **kwds)
|
||||
|
||||
def gg(t, *args, **kwds):
|
||||
return self.g(scale * t + offset, *args, **kwds)
|
||||
|
||||
def dgg(t, *args, **kwds):
|
||||
return scale * self.dg(scale * t + offset, *args, **kwds)
|
||||
dg_a = abs(omega * dgg(-1, *args, **kwds))
|
||||
dg_b = abs(omega * dgg(1, *args, **kwds))
|
||||
g_a = abs(omega * gg(-1, *args, **kwds))
|
||||
g_b = abs(omega * gg(1, *args, **kwds))
|
||||
delta, alpha = min(dg_a, dg_b), min(g_a, g_b)
|
||||
|
||||
betam = delta # * scale
|
||||
if self.endpoints:
|
||||
if delta < 10 or alpha <= 10 or s > 1:
|
||||
points = chebyshev_extrema
|
||||
else:
|
||||
points = adaptive_levin_points
|
||||
elif delta < 10 or alpha <= 10 or s > 1:
|
||||
points = chebyshev_roots
|
||||
else:
|
||||
points = open_levin_points # tanh_sinh_open_nodes
|
||||
|
||||
m = self._get_num_points(s, prec, betam)
|
||||
abseps = 10 * 10.0**-prec
|
||||
num_collocation_point_list = m * 2**np.arange(1, 5) + 1
|
||||
basis = self.basis
|
||||
|
||||
q_val = 1e+300
|
||||
num_function_evaluations = 0
|
||||
n = 0
|
||||
for num_collocation_points in num_collocation_point_list:
|
||||
n_old = n
|
||||
q_old = q_val
|
||||
x = points(num_collocation_points, betam)
|
||||
n = len(x)
|
||||
if n > n_old:
|
||||
q_val = self._a_levin(omega, ff, gg, dgg, x, s, basis, *args,
|
||||
**kwds)
|
||||
num_function_evaluations += n
|
||||
err = np.abs(q_val - q_old)
|
||||
if err <= abseps:
|
||||
break
|
||||
info = self.info(err, num_function_evaluations)
|
||||
return q_val, info
|
||||
|
||||
|
||||
class EvansWebster(AdaptiveLevin):
|
||||
"""Return integral for the Evans Webster method"""
|
||||
|
||||
def __init__(self, f, g, dg=None, a=-1, b=1, basis=chebyshev_basis, s=8,
|
||||
precision=10, endpoints=False, full_output=False):
|
||||
super(EvansWebster,
|
||||
self).__init__(f, g, dg=dg, a=a, b=b, basis=basis, s=s,
|
||||
precision=precision, endpoints=endpoints,
|
||||
full_output=full_output)
|
||||
|
||||
def _a_levin(self, omega, ff, gg, dgg, x, s, basis, *args, **kwds):
|
||||
w = evans_webster_weights(omega, gg, dgg, x, basis, *args, **kwds)
|
||||
|
||||
f = Limit(ff)(x, *args, **kwds)
|
||||
return np.sum(f * w)
|
||||
|
||||
def _get_num_points(self, s, prec, betam):
|
||||
return 8 if s > 1 else int(prec / max(np.log10(betam + 1), 1) + 1)
|
||||
|
||||
|
||||
if __name__ == '__main__':
|
||||
tanh_sinh_nodes(16)
|
@ -1,532 +0,0 @@
|
||||
'''
|
||||
|
||||
All the software contained in this library is protected by copyright.
|
||||
Permission to use, copy, modify, and distribute this software for any
|
||||
purpose without fee is hereby granted, provided that this entire notice
|
||||
is included in all copies of any software which is or includes a copy
|
||||
or modification of this software and in all copies of the supporting
|
||||
documentation for such software.
|
||||
|
||||
THIS SOFTWARE IS BEING PROVIDED "AS IS", WITHOUT ANY EXPRESS OR IMPLIED
|
||||
WARRANTY. IN NO EVENT, NEITHER THE AUTHORS, NOR THE PUBLISHER, NOR ANY
|
||||
MEMBER OF THE EDITORIAL BOARD OF THE JOURNAL "NUMERICAL ALGORITHMS",
|
||||
NOR ITS EDITOR-IN-CHIEF, BE LIABLE FOR ANY ERROR IN THE SOFTWARE, ANY
|
||||
MISUSE OF IT OR ANY DAMAGE ARISING OUT OF ITS USE. THE ENTIRE RISK OF
|
||||
USING THE SOFTWARE LIES WITH THE PARTY DOING SO.
|
||||
|
||||
ANY USE OF THE SOFTWARE CONSTITUTES ACCEPTANCE OF THE TERMS OF THE
|
||||
ABOVE STATEMENT.
|
||||
|
||||
|
||||
AUTHORS:
|
||||
Per A Brodtkorb
|
||||
Python code Based on matlab code written by:
|
||||
|
||||
Marco Caliari
|
||||
University of Verona, Italy
|
||||
E-mail: marco.caliari@univr.it
|
||||
|
||||
Stefano de Marchi, Alvise Sommariva, Marco Vianello
|
||||
University of Padua, Italy
|
||||
E-mail: demarchi@math.unipd.it, alvise@math.unipd.it,
|
||||
marcov@math.unipd.it
|
||||
|
||||
Reference
|
||||
---------
|
||||
Padua2DM: fast interpolation and cubature at the Padua points in Matlab/Octave
|
||||
NUMERICAL ALGORITHMS, 56 (2011), PP. 45-60
|
||||
|
||||
|
||||
Padua module
|
||||
------------
|
||||
In polynomial interpolation of two variables, the Padua points are the first
|
||||
known example (and up to now the only one) of a unisolvent point set
|
||||
(that is, the interpolating polynomial is unique) with minimal growth of their
|
||||
Lebesgue constant, proven to be O(log2 n).
|
||||
This module provides all the functions needed to perform interpolation and
|
||||
cubature at the Padua points, together with the functions and the demos used
|
||||
in the paper.
|
||||
|
||||
pdint.m : main function for interpolation at the Padua points
|
||||
pdcub.m : main function for cubature at the Padua points
|
||||
pdpts.m : function for the computation of the Padua points
|
||||
padua_fit.m : function for the computation of the interpolation
|
||||
coefficients by FFT (recommended)
|
||||
pdcfsMM.m : function for the computation of the interpolation
|
||||
coefficients by matrix multiplications
|
||||
pdval.m : function for the evaluation of the interpolation
|
||||
polynomial
|
||||
pdwtsFFT.m : function for the computation of the cubature
|
||||
weights by FFT
|
||||
pdwtsMM.m : function for the computation of the cubature
|
||||
weights by matrix multiplications (recommended)
|
||||
funct.m : function containing some test functions
|
||||
demo_pdint.m : demo script for pdint
|
||||
demo_cputime_pdint.m : demo script for the computation of CPU time for
|
||||
interpolation
|
||||
demo_errors_pdint.m : demo script for the comparison of interpolation with
|
||||
coefficients computed by FFT or by matrix
|
||||
multiplications
|
||||
demo_pdcub : demo script for pdcub
|
||||
demo_cputime_pdcub.m : demo script for the computation of CPU time for
|
||||
cubature
|
||||
demo_errors_pdcub.m : demo script for the comparison of cubature with
|
||||
weights computed by FFT or by matrix multiplications
|
||||
demo_errors_pdcub_gl.m : demo script for the comparison of different cubature
|
||||
formulas
|
||||
cubature_square.m : function for the computation of some cubature
|
||||
formulas for the square
|
||||
omelyan_solovyan_rule.m : function for the computation of Omelyan-Solovyan
|
||||
cubature points and weights
|
||||
Contents.m : Contents file for Matlab
|
||||
|
||||
|
||||
'''
|
||||
from __future__ import absolute_import, division
|
||||
import numpy as np
|
||||
from numpy.fft import fft
|
||||
from wafo.dctpack import dct
|
||||
from wafo.polynomial import map_from_interval, map_to_interval
|
||||
# from scipy.fftpack.realtransforms import dct
|
||||
|
||||
|
||||
class _ExampleFunctions(object):
|
||||
'''
|
||||
Computes test function in the points (x, y)
|
||||
|
||||
Parameters
|
||||
----------
|
||||
x,y : array-like
|
||||
evaluate the function in the points (x,y)
|
||||
i : scalar int (default 0)
|
||||
defining which test function to use. Options are
|
||||
0: franke
|
||||
1: half_sphere
|
||||
2: poly_degree20
|
||||
3: exp_fun1
|
||||
4: exp_fun100
|
||||
5: cos30
|
||||
6: constant
|
||||
7: exp_xy
|
||||
8: runge
|
||||
9: abs_cubed
|
||||
10: gauss
|
||||
11: exp_inv
|
||||
|
||||
Returns
|
||||
-------
|
||||
z : array-like
|
||||
value of the function in the points (x,y)
|
||||
'''
|
||||
@staticmethod
|
||||
def franke(x, y):
|
||||
'''Franke function.
|
||||
|
||||
The value of the definite integral on the square [-1,1] x [-1,1],
|
||||
obtained using a Padua Points cubature formula of degree 500, is
|
||||
2.1547794245591083e+000 with an estimated absolute error of 8.88e-016.
|
||||
|
||||
The value of the definite integral on the square [0,1] x [0,1],
|
||||
obtained using a Padua Points cubature formula of degree 500, is
|
||||
4.06969589491556e-01 with an estimated absolute error of 8.88e-016.
|
||||
|
||||
Maple: 0.40696958949155611906
|
||||
'''
|
||||
def _exp(x, y, loc, scale, p2=2):
|
||||
return np.exp(- (x - loc[0])**2 / scale[0] - (y - loc[1])**p2 / scale[1])
|
||||
# exp = np.exp
|
||||
x9, y9 = 9. * x, 9. * y
|
||||
return (3. / 4 * _exp(x9, y9, [2, 2], [4, 4]) +
|
||||
3. / 4 * _exp(x9, y9, [-1, -1], [49, 10], p2=1) +
|
||||
1. / 2 * _exp(x9, y9, [7, 3], [4, 4]) -
|
||||
1. / 5 * _exp(x9, y9, [4, 7], [1, 1]))
|
||||
|
||||
@staticmethod
|
||||
def half_sphere(x, y):
|
||||
'''The value of the definite integral on the square [-1,1] x [-1,1],
|
||||
obtained using a Padua Points cubature formula of degree 2000, is
|
||||
3.9129044444568244e+000 with an estimated absolute error of 3.22e-010.
|
||||
'''
|
||||
return ((x - 0.5)**2 + (y - 0.5)**2)**(1. / 2)
|
||||
|
||||
@staticmethod
|
||||
def poly_degree20(x, y):
|
||||
''''Bivariate polynomial having moderate degree.
|
||||
The value of the definite integral on the square [-1,1] x
|
||||
[-1,1], up to machine precision, is 18157.16017316017 (see ref. 6).
|
||||
The value of the definite integral on the square [-1,1] x [-1,1],
|
||||
obtained using a Padua Points cubature formula of degree 500,
|
||||
is 1.8157160173160162e+004.
|
||||
|
||||
2D modification of an example by L.N.Trefethen (see ref. 7), f(x)=x^20.
|
||||
'''
|
||||
return (x + y)**20
|
||||
|
||||
@staticmethod
|
||||
def exp_fun1(x, y):
|
||||
''' The value of the definite integral on the square [-1,1] x [-1,1],
|
||||
obtained using a Padua Points cubature formula of degree 2000,
|
||||
is 2.1234596326670683e+001 with an estimated absolute error of
|
||||
7.11e-015.
|
||||
'''
|
||||
return np.exp((x - 0.5)**2 + (y - 0.5)**2)
|
||||
|
||||
@staticmethod
|
||||
def exp_fun100(x, y):
|
||||
'''The value of the definite integral on the square [-1,1] x [-1,1],
|
||||
obtained using a Padua Points cubature formula of degree 2000,
|
||||
is 3.1415926535849605e-002 with an estimated absolute error of
|
||||
3.47e-017.
|
||||
'''
|
||||
return np.exp(-100 * ((x - 0.5)**2 + (y - 0.5)**2))
|
||||
|
||||
@staticmethod
|
||||
def cos30(x, y):
|
||||
''' The value of the definite integral on the square [-1,1] x [-1,1],
|
||||
obtained using a Padua Points cubature formula of degree 500,
|
||||
is 4.3386955120336568e-003 with an estimated absolute error of
|
||||
2.95e-017.
|
||||
'''
|
||||
return np.cos(30 * (x + y))
|
||||
|
||||
@staticmethod
|
||||
def constant(x, y):
|
||||
'''Constant.
|
||||
To test interpolation and cubature at degree 0.
|
||||
The value of the definite integral on the square [-1,1] x [-1,1]
|
||||
is 4.
|
||||
'''
|
||||
return np.ones(np.shape(x + y))
|
||||
|
||||
@staticmethod
|
||||
def exp_xy(x, y):
|
||||
'''The value of the definite integral on the square [-1,1] x [-1,1]
|
||||
is up to machine precision is 5.524391382167263 (see ref. 6).
|
||||
2. The value of the definite integral on the square [-1,1] x [-1,1],
|
||||
obtained using a Padua Points cubature formula of degree 500,
|
||||
is 5.5243913821672628e+000 with an estimated absolute error of
|
||||
0.00e+000.
|
||||
2D modification of an example by L.N.Trefethen (see ref. 7),
|
||||
f(x)=exp(x).
|
||||
'''
|
||||
return np.exp(x + y)
|
||||
|
||||
@staticmethod
|
||||
def runge(x, y):
|
||||
''' Bivariate Runge function: as 1D complex function is analytic
|
||||
in a neighborhood of [-1; 1] but not throughout the complex plane.
|
||||
|
||||
The value of the definite integral on the square [-1,1] x [-1,1],
|
||||
up to machine precision, is 0.597388947274307 (see ref. 6).
|
||||
The value of the definite integral on the square [-1,1] x [-1,1],
|
||||
obtained using a Padua Points cubature formula of degree 500,
|
||||
is 5.9738894727430725e-001 with an estimated absolute error of
|
||||
0.00e+000.
|
||||
|
||||
2D modification of an example by L.N.Trefethen (see ref. 7),
|
||||
f(x)=1/(1+16*x^2).
|
||||
'''
|
||||
return 1. / (1 + 16 * (x**2 + y**2))
|
||||
|
||||
@staticmethod
|
||||
def abs_cubed(x, y):
|
||||
'''Low regular function.
|
||||
The value of the definite integral on the square [-1,1] x [-1,1],
|
||||
up to machine precision, is 2.508723139534059 (see ref. 6).
|
||||
The value of the definite integral on the square [-1,1] x [-1,1],
|
||||
obtained using a Padua Points cubature formula of degree 500,
|
||||
is 2.5087231395340579e+000 with an estimated absolute error of
|
||||
0.00e+000.
|
||||
|
||||
2D modification of an example by L.N.Trefethen (see ref. 7),
|
||||
f(x)=abs(x)^3.
|
||||
'''
|
||||
return (x**2 + y**2)**(3 / 2)
|
||||
|
||||
@staticmethod
|
||||
def gauss(x, y):
|
||||
'''Bivariate gaussian: smooth function.
|
||||
The value of the definite integral on the square [-1,1] x [-1,1],
|
||||
up to machine precision, is 2.230985141404135 (see ref. 6).
|
||||
The value of the definite integral on the square [-1,1] x [-1,1],
|
||||
obtained using a Padua Points cubature formula of degree 500,
|
||||
is 2.2309851414041333e+000 with an estimated absolute error of
|
||||
2.66e-015.
|
||||
|
||||
2D modification of an example by L.N.Trefethen (see ref. 7),
|
||||
f(x)=exp(-x^2).
|
||||
'''
|
||||
return np.exp(-x**2 - y**2)
|
||||
|
||||
@staticmethod
|
||||
def exp_inv(x, y):
|
||||
'''Bivariate example stemming from a 1D C-infinity function.
|
||||
The value of the definite integral on the square [-1,1] x [-1,1],
|
||||
up to machine precision, is 0.853358758654305 (see ref. 6).
|
||||
The value of the definite integral on the square [-1,1] x [-1,1],
|
||||
obtained using a Padua Points cubature formula of degree 2000,
|
||||
is 8.5335875865430544e-001 with an estimated absolute error of
|
||||
3.11e-015.
|
||||
|
||||
2D modification of an example by L.N.Trefethen (see ref. 7),
|
||||
f(x)=exp(-1/x^2).
|
||||
'''
|
||||
arg_z = (x**2 + y**2)
|
||||
# Avoid cases in which "arg_z=0", setting only in those instances
|
||||
# "arg_z=eps".
|
||||
arg_z = arg_z + (1 - np.abs(np.sign(arg_z))) * 1.e-100
|
||||
arg_z = 1. / arg_z
|
||||
return np.exp(-arg_z)
|
||||
|
||||
def __call__(self, x, y, i=0):
|
||||
s = self
|
||||
test_function = [s.franke, s.half_sphere, s.poly_degree20, s.exp_fun1,
|
||||
s.exp_fun100, s.cos30, s.constant, s.exp_xy, s.runge,
|
||||
s.abs_cubed, s.gauss, s.exp_inv]
|
||||
return test_function[i](x, y)
|
||||
|
||||
|
||||
example_functions = _ExampleFunctions()
|
||||
|
||||
|
||||
def _find_m(n):
|
||||
ix = np.r_[1:(n + 1) * (n + 2):2]
|
||||
if np.mod(n, 2) == 0:
|
||||
n2 = n // 2
|
||||
offset = np.array([[0, 1] * n2 + [0, ]] * (n2 + 1))
|
||||
ix = ix - offset.ravel(order='F')
|
||||
return ix
|
||||
|
||||
|
||||
def padua_points(n, domain=(-1, 1, -1, 1)):
|
||||
''' Return Padua points
|
||||
|
||||
Parameters
|
||||
----------
|
||||
n : scalar integer
|
||||
interpolation degree
|
||||
domain : vector [a,b,c,d]
|
||||
defining the rectangle [a,b] x [c,d]. (default domain = (-1,1,-1,1))
|
||||
|
||||
Returns
|
||||
-------
|
||||
pad : array of shape (2 x (n+1)*(n+2)/2) such that
|
||||
(pad[0,:], pad[1,: ]) defines the Padua points in the domain
|
||||
rectangle [a,b] x [c,d].
|
||||
or
|
||||
X1,Y1,X2,Y2 : arrays
|
||||
Two subgrids X1,Y1 and X2,Y2 defining the Padua points
|
||||
-------------------------------------------------------------------------------
|
||||
'''
|
||||
a, b, c, d = domain
|
||||
t0 = [np.pi] if n == 0 else np.linspace(0, np.pi, n + 1)
|
||||
t1 = np.linspace(0, np.pi, n + 2)
|
||||
zn = map_to_interval(np.cos(t0), a, b)
|
||||
zn1 = map_to_interval(np.cos(t1), c, d)
|
||||
|
||||
Pad1, Pad2 = np.meshgrid(zn, zn1)
|
||||
ix = _find_m(n)
|
||||
return np.vstack((Pad1.ravel(order='F')[ix],
|
||||
Pad2.ravel(order='F')[ix]))
|
||||
|
||||
|
||||
def error_estimate(C0f):
|
||||
''' Return interpolation error estimate from Padua coefficients
|
||||
'''
|
||||
n = C0f.shape[1]
|
||||
C0f2 = np.fliplr(C0f)
|
||||
errest = sum(np.abs(np.diag(C0f2)))
|
||||
if (n >= 1):
|
||||
errest = errest + sum(np.abs(np.diag(C0f2, -1)))
|
||||
if (n >= 2):
|
||||
errest = errest + sum(np.abs(np.diag(C0f2, -2)))
|
||||
return 2 * errest
|
||||
|
||||
|
||||
def padua_fit(Pad, fun, *args):
|
||||
'''
|
||||
Computes the Chebyshevs coefficients
|
||||
|
||||
so that f(x, y) can be approximated by:
|
||||
|
||||
f(x, y) = sum cjk*Tjk(x, y)
|
||||
|
||||
Parameters
|
||||
----------
|
||||
Pad : array-like
|
||||
Padua points, as computed with padua_points function.
|
||||
fun : function to be interpolated in the form
|
||||
fun(x, y, *args), where *args are optional arguments for fun.
|
||||
|
||||
Returns
|
||||
-------
|
||||
coefficents: coefficient matrix
|
||||
abs_err : interpolation error estimate
|
||||
|
||||
Example
|
||||
------
|
||||
>>> import numpy as np
|
||||
>>> import wafo.padua as wp
|
||||
>>> domain = [0, 1, 0, 1]
|
||||
>>> a, b, c, d = domain
|
||||
>>> points = wp.padua_points(21, domain)
|
||||
>>> C0f, abs_error = wp.padua_fit(points, wp.example_functions.franke)
|
||||
>>> x1 = np.linspace(a, b, 100)
|
||||
>>> x2 = np.linspace(c, d, 101)
|
||||
>>> val = wp.padua_val(x1, x2, C0f, domain, use_meshgrid=True)
|
||||
>>> X1, X2 = np.meshgrid(x1, x2)
|
||||
>>> true_val = wp.example_functions.franke(X1, X2)
|
||||
|
||||
>>> np.allclose(val, true_val, atol=10*abs_error)
|
||||
True
|
||||
>>> np.allclose(np.abs(val-true_val).max(), 0.0073174614275738296)
|
||||
True
|
||||
>>> np.allclose(abs_error, 0.0022486904061664046)
|
||||
True
|
||||
|
||||
import matplotlib.pyplot as plt
|
||||
plt.contour(x1, x2, val)
|
||||
|
||||
'''
|
||||
|
||||
N = np.shape(Pad)[1]
|
||||
# recover the degree n from N = (n+1)(n+2)/2
|
||||
n = int(round(-3 + np.sqrt(1 + 8 * N)) / 2)
|
||||
C0f = fun(Pad[0], Pad[1], *args)
|
||||
if (n > 0):
|
||||
ix = _find_m(n)
|
||||
GfT = np.zeros((n + 2) * (n + 1))
|
||||
GfT[ix] = C0f * 2 / (n * (n + 1))
|
||||
GfT = GfT.reshape(n + 1, n + 2)
|
||||
GfT = GfT.T
|
||||
GfT[0] = GfT[0] / 2
|
||||
GfT[n + 1] = GfT[n + 1] / 2
|
||||
GfT[:, 0] = GfT[:, 0] / 2
|
||||
GfT[:, n] = GfT[:, n] / 2
|
||||
Gf = GfT.T
|
||||
# compute the interpolation coefficient matrix C0f by FFT
|
||||
Gfhat = np.real(fft(Gf, 2 * n, axis=0))
|
||||
Gfhathat = np.real(fft(Gfhat[:n + 1, :], 2 * (n + 1), axis=1))
|
||||
C0f = 2 * Gfhathat[:, 0:n + 1]
|
||||
C0f[0] = C0f[0, :] / np.sqrt(2)
|
||||
C0f[:, 0] = C0f[:, 0] / np.sqrt(2)
|
||||
C0f = np.fliplr(np.triu(np.fliplr(C0f)))
|
||||
C0f[n] = C0f[n] / 2
|
||||
|
||||
return C0f, error_estimate(C0f)
|
||||
|
||||
|
||||
def paduavals2coefs(f):
|
||||
m = len(f)
|
||||
n = int(round(-1.5 + np.sqrt(.25 + 2 * m)))
|
||||
x = padua_points(n)
|
||||
idx = _find_m(n)
|
||||
w = 0 * x[0] + 1. / (n * (n + 1))
|
||||
idx1 = np.all(np.abs(x) == 1, axis=0)
|
||||
w[idx1] = .5 * w[idx1]
|
||||
idx2 = np.all(np.abs(x) != 1, axis=0)
|
||||
w[idx2] = 2 * w[idx2]
|
||||
|
||||
G = np.zeros(idx.max() + 1)
|
||||
G[idx] = 4 * w * f
|
||||
|
||||
use_dct = 100 < n
|
||||
if use_dct:
|
||||
C = np.rot90(dct(dct(G.T).T)) # , axis=1)
|
||||
else:
|
||||
t1 = np.r_[0:n + 1].reshape(-1, 1)
|
||||
Tn1 = np.cos(t1 * t1.T * np.pi / n)
|
||||
t2 = np.r_[0:n + 2].reshape(-1, 1)
|
||||
Tn2 = np.cos(t2 * t2.T * np.pi / (n + 1))
|
||||
C = np.dot(Tn2, np.dot(G, Tn1))
|
||||
|
||||
C[0] = .5 * C[0]
|
||||
C[:, 1] = .5 * C[:, 1]
|
||||
C[0, -1] = .5 * C[0, -1]
|
||||
del C[-1]
|
||||
|
||||
# Take upper-left triangular part:
|
||||
return np.fliplr(np.triu(np.fliplr(C)))
|
||||
# C = triu(C(:,end:-1:1));
|
||||
# C = C(:,end:-1:1);
|
||||
|
||||
|
||||
# TODO: padua_fit2 does not work correctly yet.
|
||||
def padua_fit2(Pad, fun, *args):
|
||||
# N = np.shape(Pad)[1]
|
||||
# recover the degree n from N = (n+1)(n+2)/2
|
||||
# _n = int(round(-3 + np.sqrt(1 + 8 * N)) / 2)
|
||||
C0f = fun(Pad[0], Pad[1], *args)
|
||||
return paduavals2coefs(C0f)
|
||||
|
||||
|
||||
def _compute_moments(n):
|
||||
k = np.r_[0:n:2]
|
||||
mom = 2 * np.sqrt(2) / (1 - k ** 2)
|
||||
mom[0] = 2
|
||||
return mom
|
||||
|
||||
|
||||
def padua_cubature(coefficients, domain=(-1, 1, -1, 1)):
|
||||
'''
|
||||
Compute the integral through the coefficient matrix.
|
||||
'''
|
||||
n = coefficients.shape[1]
|
||||
mom = _compute_moments(n)
|
||||
M1, M2 = np.meshgrid(mom, mom)
|
||||
M = M1 * M2
|
||||
C0fM = coefficients[0:n:2, 0:n:2] * M
|
||||
a, b, c, d = domain
|
||||
integral = (b - a) * (d - c) * C0fM.sum() / 4
|
||||
return integral
|
||||
|
||||
|
||||
def padua_val(X, Y, coefficients, domain=(-1, 1, -1, 1), use_meshgrid=False):
|
||||
'''
|
||||
Evaluate polynomial in padua form at X, Y.
|
||||
|
||||
Evaluate the interpolation polynomial defined through its coefficient
|
||||
matrix coefficients at the target points X(:,1),X(:,2) or at the
|
||||
meshgrid(X1,X2)
|
||||
|
||||
Parameters
|
||||
----------
|
||||
X, Y: array-like
|
||||
evaluation points.
|
||||
coefficients : array-like
|
||||
coefficient matrix
|
||||
domain : a vector [a,b,c,d]
|
||||
defining the rectangle [a,b] x [c,d]
|
||||
use_meshgrid: bool
|
||||
If True interpolate at the points meshgrid(X, Y)
|
||||
|
||||
Returns
|
||||
-------
|
||||
fxy : array-like
|
||||
evaluation of the interpolation polynomial at the target points
|
||||
'''
|
||||
def transform(tn, x, a, b):
|
||||
xn = map_from_interval(x, a, b).clip(min=-1, max=1).reshape(1, -1)
|
||||
tx = np.cos(tn * np.arccos(xn)) * np.sqrt(2)
|
||||
tx[0] = 1
|
||||
return tx
|
||||
|
||||
X, Y = np.atleast_1d(X, Y)
|
||||
original_shape = X.shape
|
||||
a, b, c, d = domain
|
||||
n = np.shape(coefficients)[1]
|
||||
|
||||
tn = np.r_[0:n][:, None]
|
||||
tx1 = transform(tn, X.ravel(), a, b)
|
||||
tx2 = transform(tn, Y.ravel(), c, d)
|
||||
|
||||
if use_meshgrid: # eval on meshgrid points
|
||||
return np.dot(tx1.T, np.dot(coefficients, tx2)).T
|
||||
# scattered points
|
||||
val = np.sum(np.dot(tx1.T, coefficients) * tx2.T, axis=1)
|
||||
return val.reshape(original_shape)
|
||||
|
||||
|
||||
if __name__ == '__main__':
|
||||
from wafo.testing import test_docstrings
|
||||
test_docstrings(__file__)
|
@ -1,18 +0,0 @@
|
||||
# -*- coding: utf-8 -*-
|
||||
"""
|
||||
Created on Sun Oct 25 14:55:34 2015
|
||||
|
||||
@author: dave
|
||||
"""
|
||||
|
||||
|
||||
def configuration(parent_package='', top_path=None):
|
||||
from numpy.distutils.misc_util import Configuration
|
||||
config = Configuration('wafo', parent_package, top_path)
|
||||
config.add_subpackage('source')
|
||||
config.make_config_py()
|
||||
return config
|
||||
|
||||
if __name__ == "__main__":
|
||||
from numpy.distutils.core import setup
|
||||
setup(**configuration(top_path='').todict())
|
@ -1,2 +0,0 @@
|
||||
from wafo.sg_filter._core import * # pylint: disable=wildcard-import
|
||||
from wafo.sg_filter import demos
|
File diff suppressed because one or more lines are too long
@ -1,488 +0,0 @@
|
||||
import numpy as np
|
||||
from scipy.sparse.linalg import expm
|
||||
from scipy.signal import medfilt
|
||||
from wafo.plotbackend import plotbackend as plt
|
||||
from wafo.sg_filter._core import (SavitzkyGolay, smoothn, Kalman,
|
||||
HodrickPrescott, HampelFilter)
|
||||
|
||||
|
||||
def demo_savitzky_on_noisy_chirp():
|
||||
"""
|
||||
Example
|
||||
-------
|
||||
>>> demo_savitzky_on_noisy_chirp()
|
||||
|
||||
>>> plt.close()
|
||||
"""
|
||||
plt.figure(figsize=(7, 12))
|
||||
|
||||
# generate chirp signal
|
||||
tvec = np.arange(0, 6.28, .02)
|
||||
true_signal = np.sin(tvec * (2.0 + tvec))
|
||||
true_d_signal = (2 + tvec) * np.cos(tvec * (2.0 + tvec))
|
||||
|
||||
# add noise to signal
|
||||
noise = np.random.normal(size=true_signal.shape)
|
||||
signal = true_signal + .15 * noise
|
||||
|
||||
# plot signal
|
||||
plt.subplot(311)
|
||||
plt.plot(signal)
|
||||
plt.title('signal')
|
||||
|
||||
# smooth and plot signal
|
||||
plt.subplot(312)
|
||||
savgol = SavitzkyGolay(n=8, degree=4)
|
||||
s_signal = savgol.smooth(signal)
|
||||
s2 = smoothn(signal, robust=True)
|
||||
plt.plot(s_signal)
|
||||
plt.plot(s2)
|
||||
plt.plot(true_signal, 'r--')
|
||||
plt.title('smoothed signal')
|
||||
|
||||
# smooth derivative of signal and plot it
|
||||
plt.subplot(313)
|
||||
savgol1 = SavitzkyGolay(n=8, degree=1, diff_order=1)
|
||||
|
||||
dt = tvec[1] - tvec[0]
|
||||
d_signal = savgol1.smooth(signal) / dt
|
||||
|
||||
plt.plot(d_signal)
|
||||
plt.plot(true_d_signal, 'r--')
|
||||
plt.title('smoothed derivative of signal')
|
||||
|
||||
|
||||
def demo_kalman_voltimeter():
|
||||
"""
|
||||
Example
|
||||
-------
|
||||
>>> demo_kalman_voltimeter()
|
||||
|
||||
>>> plt.close()
|
||||
"""
|
||||
V0 = 12
|
||||
h = np.atleast_2d(1) # voltimeter measure the voltage itself
|
||||
q = 1e-9 # variance of process noise as the car operates
|
||||
r = 0.05 ** 2 # variance of measurement error
|
||||
b = 0 # no system input
|
||||
u = 0 # no system input
|
||||
filt = Kalman(R=r, A=1, Q=q, H=h, B=b)
|
||||
|
||||
# Generate random voltages and watch the filter operate.
|
||||
n = 50
|
||||
truth = np.random.randn(n) * np.sqrt(q) + V0
|
||||
z = truth + np.random.randn(n) * np.sqrt(r) # measurement
|
||||
x = np.zeros(n)
|
||||
|
||||
for i, zi in enumerate(z):
|
||||
x[i] = filt(zi, u) # perform a Kalman filter iteration
|
||||
|
||||
_hz = plt.plot(z, 'r.', label='observations')
|
||||
# a-posteriori state estimates:
|
||||
_hx = plt.plot(x, 'b-', label='Kalman output')
|
||||
_ht = plt.plot(truth, 'g-', label='true voltage')
|
||||
plt.legend()
|
||||
plt.title('Automobile Voltimeter Example')
|
||||
|
||||
|
||||
def lti_disc(F, L=None, Q=None, dt=1):
|
||||
"""LTI_DISC Discretize LTI ODE with Gaussian Noise.
|
||||
|
||||
Syntax:
|
||||
[A,Q] = lti_disc(F,L,Qc,dt)
|
||||
|
||||
In:
|
||||
F - NxN Feedback matrix
|
||||
L - NxL Noise effect matrix (optional, default identity)
|
||||
Qc - LxL Diagonal Spectral Density (optional, default zeros)
|
||||
dt - Time Step (optional, default 1)
|
||||
|
||||
Out:
|
||||
A - Transition matrix
|
||||
Q - Discrete Process Covariance
|
||||
|
||||
Description:
|
||||
Discretize LTI ODE with Gaussian Noise. The original
|
||||
ODE model is in form
|
||||
|
||||
dx/dt = F x + L w, w ~ N(0,Qc)
|
||||
|
||||
Result of discretization is the model
|
||||
|
||||
x[k] = A x[k-1] + q, q ~ N(0,Q)
|
||||
|
||||
Which can be used for integrating the model
|
||||
exactly over time steps, which are multiples
|
||||
of dt.
|
||||
|
||||
"""
|
||||
n = np.shape(F)[0]
|
||||
if L is None:
|
||||
L = np.eye(n)
|
||||
|
||||
if Q is None:
|
||||
Q = np.zeros((n, n))
|
||||
# Closed form integration of transition matrix
|
||||
A = expm(F * dt)
|
||||
|
||||
# Closed form integration of covariance
|
||||
# by matrix fraction decomposition
|
||||
|
||||
Phi = np.vstack((np.hstack((F, np.dot(np.dot(L, Q), L.T))),
|
||||
np.hstack((np.zeros((n, n)), -F.T))))
|
||||
AB = np.dot(expm(Phi * dt), np.vstack((np.zeros((n, n)), np.eye(n))))
|
||||
# Q = AB[:n, :] / AB[n:(2 * n), :]
|
||||
Q = np.linalg.solve(AB[n:(2 * n), :].T, AB[:n, :].T)
|
||||
return A, Q
|
||||
|
||||
|
||||
def demo_kalman_sine():
|
||||
"""Kalman Filter demonstration with sine signal.
|
||||
|
||||
Example
|
||||
-------
|
||||
>>> demo_kalman_sine()
|
||||
|
||||
>>> plt.close()
|
||||
"""
|
||||
sd = 0.5
|
||||
dt = 0.1
|
||||
w = 1
|
||||
T = np.arange(0, 30 + dt / 2, dt)
|
||||
n = len(T)
|
||||
X = 3 * np.sin(w * T)
|
||||
Y = X + sd * np.random.randn(n)
|
||||
|
||||
''' Initialize KF to values
|
||||
x = 0
|
||||
dx/dt = 0
|
||||
with great uncertainty in derivative
|
||||
'''
|
||||
M = np.zeros((2, 1))
|
||||
P = np.diag([0.1, 2])
|
||||
R = sd ** 2
|
||||
H = np.atleast_2d([1, 0])
|
||||
q = 0.1
|
||||
F = np.atleast_2d([[0, 1],
|
||||
[0, 0]])
|
||||
A, Q = lti_disc(F, L=None, Q=np.diag([0, q]), dt=dt)
|
||||
|
||||
# Track and animate
|
||||
m = M.shape[0]
|
||||
_MM = np.zeros((m, n))
|
||||
_PP = np.zeros((m, m, n))
|
||||
'''In this demonstration we estimate a stationary sine signal from noisy
|
||||
measurements by using the classical Kalman filter.'
|
||||
'''
|
||||
filt = Kalman(R=R, x=M, P=P, A=A, Q=Q, H=H, B=0)
|
||||
|
||||
# Generate random voltages and watch the filter operate.
|
||||
# n = 50
|
||||
# truth = np.random.randn(n) * np.sqrt(q) + V0
|
||||
# z = truth + np.random.randn(n) * np.sqrt(r) # measurement
|
||||
truth = X
|
||||
z = Y
|
||||
x = np.zeros((n, m))
|
||||
|
||||
for i, zi in enumerate(z):
|
||||
x[i] = np.ravel(filt(zi, u=0))
|
||||
|
||||
_hz = plt.plot(z, 'r.', label='observations')
|
||||
# a-posteriori state estimates:
|
||||
_hx = plt.plot(x[:, 0], 'b-', label='Kalman output')
|
||||
_ht = plt.plot(truth, 'g-', label='true voltage')
|
||||
plt.legend()
|
||||
plt.title('Automobile Voltimeter Example')
|
||||
|
||||
|
||||
# for k in range(m):
|
||||
# [M,P] = kf_predict(M,P,A,Q);
|
||||
# [M,P] = kf_update(M,P,Y(k),H,R);
|
||||
#
|
||||
# MM(:,k) = M;
|
||||
# PP(:,:,k) = P;
|
||||
#
|
||||
# %
|
||||
# % Animate
|
||||
# %
|
||||
# if rem(k,10)==1
|
||||
# plot(T,X,'b--',...
|
||||
# T,Y,'ro',...
|
||||
# T(k),M(1),'k*',...
|
||||
# T(1:k),MM(1,1:k),'k-');
|
||||
# legend('Real signal','Measurements','Latest estimate',
|
||||
# 'Filtered estimate')
|
||||
# title('Estimating a noisy sine signal with Kalman filter.');
|
||||
# drawnow;
|
||||
#
|
||||
# pause;
|
||||
# end
|
||||
# end
|
||||
#
|
||||
# clc;
|
||||
# disp('In this demonstration we estimate a stationary sine signal '
|
||||
# 'from noisy measurements by using the classical Kalman filter.');
|
||||
# disp(' ');
|
||||
# disp('The filtering results are now displayed sequantially for 10 time '
|
||||
# 'step at a time.');
|
||||
# disp(' ');
|
||||
# disp('<push any key to see the filtered and smoothed results together>')
|
||||
# pause;
|
||||
# %
|
||||
# % Apply Kalman smoother
|
||||
# %
|
||||
# SM = rts_smooth(MM,PP,A,Q);
|
||||
# plot(T,X,'b--',...
|
||||
# T,MM(1,:),'k-',...
|
||||
# T,SM(1,:),'r-');
|
||||
# legend('Real signal','Filtered estimate','Smoothed estimate')
|
||||
# title('Filtered and smoothed estimate of the original signal');
|
||||
#
|
||||
# clc;
|
||||
# disp('The filtered and smoothed estimates of the signal are now '
|
||||
# 'displayed.')
|
||||
# disp(' ');
|
||||
# disp('RMS errors:');
|
||||
# %
|
||||
# % Errors
|
||||
# %
|
||||
# fprintf('KF = %.3f\nRTS = %.3f\n',...
|
||||
# sqrt(mean((MM(1,:)-X(1,:)).^2)),...
|
||||
# sqrt(mean((SM(1,:)-X(1,:)).^2)));
|
||||
|
||||
|
||||
def demo_hampel():
|
||||
"""
|
||||
Example
|
||||
-------
|
||||
>>> demo_hampel()
|
||||
|
||||
>>> plt.close()
|
||||
"""
|
||||
randint = np.random.randint
|
||||
Y = 5000 + np.random.randn(1000)
|
||||
outliers = randint(0, 1000, size=(10,))
|
||||
Y[outliers] = Y[outliers] + randint(1000, size=(10,))
|
||||
YY, res = HampelFilter(dx=3, t=3, fulloutput=True)(Y)
|
||||
YY1, res1 = HampelFilter(dx=1, t=3, adaptive=0.1, fulloutput=True)(Y)
|
||||
YY2, res2 = HampelFilter(dx=3, t=0, fulloutput=True)(Y) # median
|
||||
plt.figure(1)
|
||||
plot_hampel(Y, YY, res)
|
||||
plt.title('Standard HampelFilter')
|
||||
plt.figure(2)
|
||||
plot_hampel(Y, YY1, res1)
|
||||
plt.title('Adaptive HampelFilter')
|
||||
plt.figure(3)
|
||||
plot_hampel(Y, YY2, res2)
|
||||
plt.title('Median filter')
|
||||
|
||||
|
||||
def plot_hampel(Y, YY, res):
|
||||
X = np.arange(len(YY))
|
||||
plt.plot(X, Y, 'b.') # Original Data
|
||||
plt.plot(X, YY, 'r') # Hampel Filtered Data
|
||||
plt.plot(X, res['Y0'], 'b--') # Nominal Data
|
||||
plt.plot(X, res['LB'], 'r--') # Lower Bounds on Hampel Filter
|
||||
plt.plot(X, res['UB'], 'r--') # Upper Bounds on Hampel Filter
|
||||
i = res['outliers']
|
||||
plt.plot(X[i], Y[i], 'ks') # Identified Outliers
|
||||
|
||||
|
||||
def demo_tide_filter():
|
||||
"""
|
||||
Example
|
||||
-------
|
||||
>>> demo_tide_filter()
|
||||
|
||||
>>> plt.close()
|
||||
"""
|
||||
# import statsmodels.api as sa
|
||||
import wafo.spectrum.models as sm
|
||||
sd = 10
|
||||
Sj = sm.Jonswap(Hm0=4. * sd)
|
||||
S = Sj.tospecdata()
|
||||
|
||||
q = (0.1 * sd) ** 2 # variance of process noise s the car operates
|
||||
r = (100 * sd) ** 2 # variance of measurement error
|
||||
b = 0 # no system input
|
||||
u = 0 # no system input
|
||||
|
||||
from scipy.signal import butter, filtfilt, lfilter_zi # lfilter,
|
||||
freq_tide = 1. / (12 * 60 * 60)
|
||||
freq_wave = 1. / 10
|
||||
freq_filt = freq_wave / 10
|
||||
dt = 1.
|
||||
freq = 1. / dt
|
||||
fn = (freq / 2)
|
||||
|
||||
P = 10 * np.diag([1, 0.01])
|
||||
R = r
|
||||
H = np.atleast_2d([1, 0])
|
||||
|
||||
F = np.atleast_2d([[0, 1],
|
||||
[0, 0]])
|
||||
A, Q = lti_disc(F, L=None, Q=np.diag([0, q]), dt=dt)
|
||||
|
||||
t = np.arange(0, 60 * 12, 1. / freq)
|
||||
w = 2 * np.pi * freq # 1 Hz
|
||||
tide = 100 * np.sin(freq_tide * w * t + 2 * np.pi / 4) + 100
|
||||
y = tide + S.sim(len(t), dt=1. / freq)[:, 1].ravel()
|
||||
# lowess = sa.nonparametric.lowess
|
||||
# y2 = lowess(y, t, frac=0.5)[:,1]
|
||||
|
||||
filt = Kalman(R=R, x=np.array([[tide[0]], [0]]), P=P, A=A, Q=Q, H=H, B=b)
|
||||
filt2 = Kalman(R=R, x=np.array([[tide[0]], [0]]), P=P, A=A, Q=Q, H=H, B=b)
|
||||
# y = tide + 0.5 * np.sin(freq_wave * w * t)
|
||||
# Butterworth filter
|
||||
b, a = butter(9, (freq_filt / fn), btype='low')
|
||||
# y2 = [lowess(y[max(i-60,0):i + 1], t[max(i-60,0):i + 1], frac=.3)[-1,1]
|
||||
# for i in range(len(y))]
|
||||
# y2 = [lfilter(b, a, y[:i + 1])[i] for i in range(len(y))]
|
||||
# y3 = filtfilt(b, a, y[:16]).tolist() + [filtfilt(b, a, y[:i + 1])[i]
|
||||
# for i in range(16, len(y))]
|
||||
# y0 = medfilt(y, 41)
|
||||
_zi = lfilter_zi(b, a)
|
||||
# y2 = lfilter(b, a, y)#, zi=y[0]*zi) # standard filter
|
||||
y3 = filtfilt(b, a, y) # filter with phase shift correction
|
||||
y4 = []
|
||||
y5 = []
|
||||
for _i, j in enumerate(y):
|
||||
tmp = np.ravel(filt(j, u=u))
|
||||
tmp = np.ravel(filt2(tmp[0], u=u))
|
||||
# if i==0:
|
||||
# print(filt.x)
|
||||
# print(filt2.x)
|
||||
y4.append(tmp[0])
|
||||
y5.append(tmp[1])
|
||||
_y0 = medfilt(y4, 41)
|
||||
# print(filt.P)
|
||||
# plot
|
||||
|
||||
plt.plot(t, y, 'r.-', linewidth=2, label='raw data')
|
||||
# plt.plot(t, y2, 'b.-', linewidth=2, label='lowess @ %g Hz' % freq_filt)
|
||||
# plt.plot(t, y2, 'b.-', linewidth=2, label='filter @ %g Hz' % freq_filt)
|
||||
plt.plot(t, y3, 'g.-', linewidth=2, label='filtfilt @ %g Hz' % freq_filt)
|
||||
plt.plot(t, y4, 'k.-', linewidth=2, label='kalman')
|
||||
# plt.plot(t, y5, 'k.', linewidth=2, label='kalman2')
|
||||
plt.plot(t, tide, 'y-', linewidth=2, label='True tide')
|
||||
plt.legend(frameon=False, fontsize=14)
|
||||
plt.xlabel("Time [s]")
|
||||
plt.ylabel("Amplitude")
|
||||
|
||||
|
||||
def demo_savitzky_on_exponential():
|
||||
"""
|
||||
Example
|
||||
-------
|
||||
>>> demo_savitzky_on_exponential()
|
||||
|
||||
>>> plt.close()
|
||||
"""
|
||||
t = np.linspace(-4, 4, 500)
|
||||
y = np.exp(-t ** 2) + np.random.normal(0, 0.05, np.shape(t))
|
||||
n = 11
|
||||
ysg = SavitzkyGolay(n, degree=1, diff_order=0)(y)
|
||||
plt.plot(t, y, t, ysg, '--')
|
||||
|
||||
|
||||
def demo_smoothn_on_1d_cos():
|
||||
"""
|
||||
Example
|
||||
-------
|
||||
>>> demo_smoothn_on_1d_cos()
|
||||
|
||||
>>> plt.close()
|
||||
"""
|
||||
x = np.linspace(0, 100, 2 ** 8)
|
||||
y = np.cos(x / 10) + (x / 50) ** 2 + np.random.randn(np.size(x)) / 10
|
||||
y[np.r_[70, 75, 80]] = np.array([5.5, 5, 6])
|
||||
z = smoothn(y) # Regular smoothing
|
||||
zr = smoothn(y, robust=True) # Robust smoothing
|
||||
_h0 = plt.subplot(121),
|
||||
_h = plt.plot(x, y, 'r.', x, z, 'k', linewidth=2)
|
||||
plt.title('Regular smoothing')
|
||||
plt.subplot(122)
|
||||
plt.plot(x, y, 'r.', x, zr, 'k', linewidth=2)
|
||||
plt.title('Robust smoothing')
|
||||
|
||||
|
||||
def demo_smoothn_on_2d_exp_sin():
|
||||
"""
|
||||
Example
|
||||
-------
|
||||
>>> demo_smoothn_on_2d_exp_sin()
|
||||
|
||||
>>> plt.close()
|
||||
"""
|
||||
xp = np.arange(0, 1, 0.02) # np.r_[0:1:0.02]
|
||||
[x, y] = np.meshgrid(xp, xp)
|
||||
f = np.exp(x + y) + np.sin((x - 2 * y) * 3)
|
||||
fn = f + np.random.randn(*f.shape) * 0.5
|
||||
_fs, s = smoothn(fn, fulloutput=True)
|
||||
fs2 = smoothn(fn, s=2 * s)
|
||||
_h = plt.subplot(131),
|
||||
_h = plt.contourf(xp, xp, fn)
|
||||
_h = plt.subplot(132),
|
||||
_h = plt.contourf(xp, xp, fs2)
|
||||
_h = plt.subplot(133),
|
||||
_h = plt.contourf(xp, xp, f)
|
||||
|
||||
|
||||
def _cardioid(n=1000):
|
||||
t = np.linspace(0, 2 * np.pi, n)
|
||||
x0 = 2 * np.cos(t) * (1 - np.cos(t))
|
||||
y0 = 2 * np.sin(t) * (1 - np.cos(t))
|
||||
x = x0 + np.random.randn(x0.size) * 0.1
|
||||
y = y0 + np.random.randn(y0.size) * 0.1
|
||||
return x, y, x0, y0
|
||||
|
||||
|
||||
def demo_smoothn_on_cardioid():
|
||||
"""
|
||||
Example
|
||||
-------
|
||||
>>> demo_smoothn_on_cardioid()
|
||||
|
||||
>>> plt.close()
|
||||
"""
|
||||
x, y, x0, y0 = _cardioid()
|
||||
z = smoothn(x + 1j * y, robust=False)
|
||||
plt.plot(x0, y0, 'y',
|
||||
x, y, 'r.',
|
||||
np.real(z), np.imag(z), 'k', linewidth=2)
|
||||
|
||||
|
||||
def demo_hodrick_on_cardioid():
|
||||
"""
|
||||
Example
|
||||
-------
|
||||
>>> demo_hodrick_on_cardioid()
|
||||
|
||||
>>> plt.close()
|
||||
"""
|
||||
x, y, x0, y0 = _cardioid()
|
||||
|
||||
smooth = HodrickPrescott(w=20000)
|
||||
# smooth = HampelFilter(adaptive=50)
|
||||
xs, ys = smooth(x), smooth(y)
|
||||
plt.plot(x0, y0, 'y',
|
||||
x, y, 'r.',
|
||||
xs, ys, 'k', linewidth=2)
|
||||
|
||||
|
||||
if __name__ == '__main__':
|
||||
from wafo.testing import test_docstrings
|
||||
test_docstrings(__file__)
|
||||
# demo_savitzky_on_noisy_chirp()
|
||||
# plt.show('hold') # show plot
|
||||
# demo_kalman_sine()
|
||||
# demo_tide_filter()
|
||||
# demo_hampel()
|
||||
# demo_kalman_voltimeter()
|
||||
# demo_savitzky_on_exponential()
|
||||
# plt.figure(1)
|
||||
# demo_hodrick_on_cardioid()
|
||||
# plt.figure(2)
|
||||
# # demo_smoothn_on_1d_cos()
|
||||
# demo_smoothn_on_cardioid()
|
||||
# plt.show('hold')
|
@ -1,11 +0,0 @@
|
||||
#!/usr/bin/env python
|
||||
# -*- coding: utf-8 -*-
|
||||
"""
|
||||
Dummy conftest.py for wafo.
|
||||
|
||||
If you don't know what this is for, just leave it empty.
|
||||
Read more about conftest.py under:
|
||||
https://pytest.org/latest/plugins.html
|
||||
"""
|
||||
from __future__ import print_function, absolute_import, division
|
||||
import pytest # @UnusedImport
|
File diff suppressed because one or more lines are too long
@ -1,25 +0,0 @@
|
||||
"""
|
||||
f2py c_library.pyf c_functions.c -c
|
||||
|
||||
See also http://www.scipy.org/Cookbook/CompilingExtensionsOnWindowsWithMinGW
|
||||
"""
|
||||
import os
|
||||
import sys
|
||||
from wafo.f2py_tools import f2py_call_str
|
||||
|
||||
def compile_all():
|
||||
f2py_call = f2py_call_str()
|
||||
print '=' * 75
|
||||
print 'compiling c_codes'
|
||||
print '=' * 75
|
||||
|
||||
compile_format = f2py_call + ' %s %s -c'
|
||||
|
||||
pyfs = ('c_library.pyf',)
|
||||
files = ('c_functions.c',)
|
||||
|
||||
for pyf, file_ in zip(pyfs, files):
|
||||
os.system(compile_format % (pyf, file_))
|
||||
|
||||
if __name__ == '__main__':
|
||||
compile_all()
|
@ -1,779 +0,0 @@
|
||||
#include "math.h"
|
||||
/*
|
||||
* Install gfortran and run the following to build the module on windows:
|
||||
* f2py c_library.pyf c_functions.c -c --fcompiler=gnu95 --compiler=mingw32 -lmsvcr71
|
||||
*/
|
||||
|
||||
/*
|
||||
* findrfc.c -
|
||||
*
|
||||
* Returns indices to RFC turningpoints of a vector
|
||||
* of turningpoints
|
||||
*
|
||||
* 1998 by Per Andreas Brodtkorb.
|
||||
*/
|
||||
|
||||
void findrfc(double *y1,double hmin, int *ind, int n,int *info) {
|
||||
double xminus, xplus, Tpl, Tmi, *y, Tstart;
|
||||
int i, j, ix=0, NC, iy;
|
||||
info[0] = 0;
|
||||
if (*(y1+0)> *(y1+1)){
|
||||
/* if first is a max , ignore the first max*/
|
||||
y=&(*(y1+1));
|
||||
NC=floor((n-1)/2);
|
||||
Tstart=1;
|
||||
}
|
||||
else {
|
||||
y=y1;
|
||||
NC=floor(n/2);
|
||||
Tstart=0;
|
||||
}
|
||||
|
||||
if (NC<1){
|
||||
return; /* No RFC cycles*/
|
||||
}
|
||||
|
||||
|
||||
if (( *(y+0) > *(y+1)) && ( *(y+1) > *(y+2)) ){
|
||||
info[0] = -1;
|
||||
return; /*This is not a sequence of turningpoints, exit */
|
||||
}
|
||||
if ((*(y+0) < *(y+1)) && (*(y+1)< *(y+2))){
|
||||
info[0]=-1;
|
||||
return; /*This is not a sequence of turningpoints, exit */
|
||||
}
|
||||
|
||||
|
||||
for (i=0; i<NC; i++) {
|
||||
|
||||
Tmi=Tstart+2*i;
|
||||
Tpl=Tstart+2*i+2;
|
||||
xminus=*(y+2*i);
|
||||
xplus=*(y+2*i+2);
|
||||
|
||||
if(i!=0){
|
||||
j=i-1;
|
||||
while((j>=0) && (*(y+2*j+1)<=*(y+2*i+1))){
|
||||
if( (*(y+2*j)<xminus) ){
|
||||
xminus=*(y+2*j);
|
||||
Tmi=Tstart+2*j;
|
||||
} /*if */
|
||||
j--;
|
||||
} /*while j*/
|
||||
} /*if i */
|
||||
if ( xminus >= xplus){
|
||||
if ( (*(y+2*i+1)-xminus) >= hmin){
|
||||
*(ind+ix)=Tmi;
|
||||
ix++;
|
||||
*(ind+ix)=(Tstart+2*i+1);
|
||||
ix++;
|
||||
} /*if*/
|
||||
goto L180;
|
||||
}
|
||||
|
||||
j=i+1;
|
||||
while((j<NC) ) {
|
||||
if (*(y+2*j+1) >= *(y+2*i+1)) goto L170;
|
||||
if( (*(y+2*j+2) <= xplus) ){
|
||||
xplus=*(y+2*j+2);
|
||||
Tpl=(Tstart+2*j+2);
|
||||
}/*if*/
|
||||
j++;
|
||||
} /*while*/
|
||||
|
||||
|
||||
if ( (*(y+2*i+1)-xminus) >= hmin) {
|
||||
*(ind+ix)=Tmi;
|
||||
ix++;
|
||||
*(ind+ix)=(Tstart+2*i+1);
|
||||
ix++;
|
||||
|
||||
} /*if*/
|
||||
goto L180;
|
||||
L170:
|
||||
if (xplus <= xminus ) {
|
||||
if ( (*(y+2*i+1)-xminus) >= hmin){
|
||||
*(ind+ix)=Tmi;
|
||||
ix++;
|
||||
*(ind+ix)=(Tstart+2*i+1);
|
||||
ix++;
|
||||
} /*if*/
|
||||
/*goto L180;*/
|
||||
}
|
||||
else{
|
||||
if ( (*(y+2*i+1)-xplus) >= hmin) {
|
||||
*(ind+ix)=(Tstart+2*i+1);
|
||||
ix++;
|
||||
*(ind+ix)=Tpl;
|
||||
ix++;
|
||||
} /*if*/
|
||||
} /*elseif*/
|
||||
L180:
|
||||
iy=i;
|
||||
} /* for i */
|
||||
info[0] = ix;
|
||||
return ;
|
||||
}
|
||||
|
||||
|
||||
|
||||
/*
|
||||
* findcross.c -
|
||||
*
|
||||
* Returns indices to level v crossings of argument vector
|
||||
*
|
||||
* 1998 by Per Andreas Brodtkorb. last modified 23.06-98
|
||||
*/
|
||||
|
||||
|
||||
void findcross(double *y, double v, int *ind, int n, int *info)
|
||||
{ int i,start, ix=0,dcross=0;
|
||||
start=0;
|
||||
if ( y[0]< v){
|
||||
dcross=-1; /* first is a up-crossing*/
|
||||
}
|
||||
else if ( y[0]> v){
|
||||
dcross=1; /* first is a down-crossing*/
|
||||
}
|
||||
else if ( y[0]== v){
|
||||
/* Find out what type of crossing we have next time.. */
|
||||
for (i=1; i<n; i++) {
|
||||
start=i;
|
||||
if ( y[i]< v){
|
||||
ind[ix] = i-1; /* first crossing is a down crossing*/
|
||||
ix++;
|
||||
dcross=-1; /* The next crossing is a up-crossing*/
|
||||
goto L120;
|
||||
}
|
||||
else if ( y[i]> v){
|
||||
ind[ix] = i-1; /* first crossing is a up-crossing*/
|
||||
ix++;
|
||||
dcross=1; /*The next crossing is a down-crossing*/
|
||||
goto L120;
|
||||
}
|
||||
}
|
||||
}
|
||||
L120:
|
||||
for (i=start; i<n-1; i++) {
|
||||
if (( (dcross==-1) && (y[i]<=v) && (y[i+1] > v) ) || ((dcross==1 ) && (y[i]>=v) && (y[i+1] < v) ) ) {
|
||||
|
||||
ind[ix] = i;
|
||||
ix++;
|
||||
dcross=-dcross;
|
||||
}
|
||||
}
|
||||
info[0] = ix;
|
||||
return;
|
||||
}
|
||||
|
||||
|
||||
/*
|
||||
* DISUFQ Is an internal function to spec2nlsdat
|
||||
*
|
||||
* CALL: disufq(rvec,ivec,rA,iA, w,kw,h,g,nmin,nmax,m,n)
|
||||
*
|
||||
* rvec, ivec = real and imaginary parts of the resultant (size m X n).
|
||||
* rA, iA = real and imaginary parts of the amplitudes (size m X n).
|
||||
* w = vector with angular frequencies (w>=0)
|
||||
* kw = vector with wavenumbers (kw>=0)
|
||||
* h = water depth (h >=0)
|
||||
* g = constant acceleration of gravity
|
||||
* nmin = minimum index where rA(:,nmin) and iA(:,nmin) is
|
||||
* greater than zero.
|
||||
* nmax = maximum index where rA(:,nmax) and iA(:,nmax) is
|
||||
* greater than zero.
|
||||
* m = size(rA,1),size(iA,1)
|
||||
* n = size(rA,2),size(iA,2), or size(rvec,2),size(ivec,2)
|
||||
*
|
||||
* DISUFQ returns the summation of difference frequency and sum
|
||||
* frequency effects in the vector vec = rvec +sqrt(-1)*ivec.
|
||||
* The 2'nd order contribution to the Stokes wave is then calculated by
|
||||
* a simple 1D Fourier transform, real(FFT(vec)).
|
||||
*
|
||||
* Install gfortran and run the following to build the module:
|
||||
* f2py diffsumfunq.pyf disufq1.c -c --fcompiler=gnu95 --compiler=mingw32 -lmsvcr71
|
||||
*
|
||||
* by Per Andreas Brodtkorb 15.08.2001
|
||||
* revised pab 14.03.2002, 01.05.2002 22.07.2002, oct 2008
|
||||
*/
|
||||
|
||||
void disufq(double *rvec, double *ivec,
|
||||
double *rA, double *iA,
|
||||
double *w, double *kw,
|
||||
double h, double g,
|
||||
int nmin, int nmax,
|
||||
int m, int n)
|
||||
{
|
||||
double Epij, Edij;
|
||||
double tmp1, tmp2, tmp3, tmp4, kfact;
|
||||
double w1, w2, kw1, kw2, Cg;
|
||||
double rrA, iiA, riA, irA;
|
||||
int i,jy,ix,iz1,iv1,ixi,jyi;
|
||||
//int iz2, iv2;
|
||||
//Initialize rvec and ivec to zero
|
||||
for (ix=0;ix<n*m;ix++) {
|
||||
rvec[ix] = 0.0;
|
||||
ivec[ix] = 0.0;
|
||||
}
|
||||
|
||||
// kfact is set to 2 in order to exploit the symmetry.
|
||||
// If you set kfact to 1, you must uncomment all statements
|
||||
// including the expressions: rvec[iz2], rvec[iv2], ivec[iz2] and ivec[iv2].
|
||||
|
||||
kfact = 2.0;
|
||||
if (h>10000){ /* deep water /Inifinite water depth */
|
||||
for (ix = nmin-1;ix<nmax;ix++) {
|
||||
ixi = ix*m;
|
||||
iz1 = 2*ixi;
|
||||
//iz2 = n*m-ixi;
|
||||
kw1 = kw[ix];
|
||||
Epij = kw1;
|
||||
for (i=0;i<m;i++,ixi++,iz1++) {
|
||||
rrA = rA[ixi]*rA[ixi]; ///
|
||||
iiA = iA[ixi]*iA[ixi]; ///
|
||||
riA = rA[ixi]*iA[ixi]; ///
|
||||
|
||||
/// Sum frequency effects along the diagonal
|
||||
tmp1 = kfact*(rrA-iiA)*Epij;
|
||||
tmp2 = kfact*2.0*riA*Epij;
|
||||
rvec[iz1] += tmp1;
|
||||
ivec[iz1] += tmp2;
|
||||
|
||||
//rvec[iz2] += tmp1;
|
||||
//ivec[iz2] -= tmp2;
|
||||
//iz2++;
|
||||
|
||||
/// Difference frequency effects are zero along the diagonal
|
||||
/// and are thus not contributing to the mean.
|
||||
}
|
||||
for (jy = ix+1;jy<nmax;jy++){
|
||||
kw2 = kw[jy];
|
||||
Epij = 0.5*(kw2 + kw1);
|
||||
Edij = -0.5*(kw2 - kw1);
|
||||
//printf("Edij = %f Epij = %f \n", Edij,Epij);
|
||||
|
||||
ixi = ix*m;
|
||||
jyi = jy*m;
|
||||
iz1 = ixi+jyi;
|
||||
iv1 = jyi-ixi;
|
||||
//iz2 = (n*m-iz1);
|
||||
//iv2 = (n*m-iv1);
|
||||
for (i = 0;i<m;i++,ixi++,jyi++,iz1++,iv1++) {
|
||||
|
||||
rrA = rA[ixi]*rA[jyi]; ///rrA = rA[i][ix]*rA[i][jy];
|
||||
iiA = iA[ixi]*iA[jyi]; ///iiA = iA[i][ix]*iA[i][jy];
|
||||
riA = rA[ixi]*iA[jyi]; ///riA = rA[i][ix]*iA[i][jy];
|
||||
irA = iA[ixi]*rA[jyi]; ///irA = iA[i][ix]*rA[i][jy];
|
||||
|
||||
/* Sum frequency effects */
|
||||
tmp1 = kfact*2.0*(rrA-iiA)*Epij;
|
||||
tmp2 = kfact*2.0*(riA+irA)*Epij;
|
||||
rvec[iz1] += tmp1;///rvec[i][ix+jy] += tmp1;
|
||||
ivec[iz1] += tmp2;///ivec[i][ix+jy] += tmp2;
|
||||
//rvec[iz2] += tmp1;///rvec[i][n*m-(ix+jy)] += tmp1;
|
||||
//ivec[iz2] -= tmp2;///ivec[i][n*m-(ix+jy)] -= tmp2;
|
||||
// iz2++;
|
||||
|
||||
/* Difference frequency effects */
|
||||
tmp1 = kfact*2.0*(rrA+iiA)*Edij;
|
||||
tmp2 = kfact*2.0*(riA-irA)*Edij;
|
||||
|
||||
rvec[iv1] += tmp1;///rvec[i][jy-ix] += tmp1;
|
||||
ivec[iv1] += tmp2;///ivec[i][jy-ix] += tmp2;
|
||||
|
||||
//rvec[iv2] += tmp1;///rvec[i][n*m-(jy-ix)] += tmp1;
|
||||
//ivec[iv2] -= tmp2;///ivec[i][n*m-(jy-ix)] -= tmp2;
|
||||
//iv2++;
|
||||
}
|
||||
}
|
||||
}
|
||||
}
|
||||
else{ /* Finite water depth */
|
||||
for (ix = nmin-1;ix<nmax;ix++) {
|
||||
kw1 = kw[ix];
|
||||
w1 = w[ix];
|
||||
tmp1 = tanh(kw1*h);
|
||||
/// Cg, wave group velocity
|
||||
Cg = 0.5*g*(tmp1 + kw1*h*(1.0- tmp1*tmp1))/w1; /// OK
|
||||
tmp1 = 0.5*g*(kw1/w1)*(kw1/w1);
|
||||
tmp2 = 0.5*w1*w1/g;
|
||||
tmp3 = g*kw1/(w1*Cg);
|
||||
|
||||
if (kw1*h<300.0){
|
||||
tmp4 = kw1/sinh(2.0*kw1*h);
|
||||
}
|
||||
else{ // To ensure sinh does not overflow.
|
||||
tmp4 = 0.0;
|
||||
}
|
||||
// Difference frequency effects finite water depth
|
||||
Edij = (tmp1-tmp2+tmp3)/(1.0-g*h/(Cg*Cg))-tmp4; /// OK
|
||||
|
||||
// Sum frequency effects finite water depth
|
||||
Epij = (3.0*(tmp1-tmp2)/(1.0-tmp1/kw1*tanh(2.0*kw1*h))+3.0*tmp2-tmp1); /// OK
|
||||
//printf("Edij = %f Epij = %f \n", Edij,Epij);
|
||||
|
||||
ixi = ix*m;
|
||||
iz1 = 2*ixi;
|
||||
//iz2 = n*m-ixi;
|
||||
for (i=0;i<m;i++,ixi++,iz1++) {
|
||||
|
||||
rrA = rA[ixi]*rA[ixi]; ///
|
||||
iiA = iA[ixi]*iA[ixi]; ///
|
||||
riA = rA[ixi]*iA[ixi]; ///
|
||||
|
||||
|
||||
/// Sum frequency effects along the diagonal
|
||||
rvec[iz1] += kfact*(rrA-iiA)*Epij;
|
||||
ivec[iz1] += kfact*2.0*riA*Epij;
|
||||
//rvec[iz2] += kfact*(rrA-iiA)*Epij;
|
||||
//ivec[iz2] -= kfact*2.0*riA*Epij;
|
||||
//iz2++;
|
||||
|
||||
/// Difference frequency effects along the diagonal
|
||||
/// are only contributing to the mean
|
||||
rvec[i] += 2.0*(rrA+iiA)*Edij;
|
||||
}
|
||||
for (jy = ix+1;jy<nmax;jy++) {
|
||||
// w1 = w[ix];
|
||||
// kw1 = kw[ix];
|
||||
w2 = w[jy];
|
||||
kw2 = kw[jy];
|
||||
tmp1 = g*(kw1/w1)*(kw2/w2);
|
||||
tmp2 = 0.5/g*(w1*w1+w2*w2+w1*w2);
|
||||
tmp3 = 0.5*g*(w1*kw2*kw2+w2*kw1*kw1)/(w1*w2*(w1+w2));
|
||||
tmp4 = (1-g*(kw1+kw2)/(w1+w2)/(w1+w2)*tanh((kw1+kw2)*h));
|
||||
Epij = (tmp1-tmp2+tmp3)/tmp4+tmp2-0.5*tmp1; /* OK */
|
||||
|
||||
tmp2 = 0.5/g*(w1*w1+w2*w2-w1*w2); /*OK*/
|
||||
tmp3 = -0.5*g*(w1*kw2*kw2-w2*kw1*kw1)/(w1*w2*(w1-w2));
|
||||
tmp4 = (1.0-g*(kw1-kw2)/(w1-w2)/(w1-w2)*tanh((kw1-kw2)*h));
|
||||
Edij = (tmp1-tmp2+tmp3)/tmp4+tmp2-0.5*tmp1; /* OK */
|
||||
//printf("Edij = %f Epij = %f \n", Edij,Epij);
|
||||
|
||||
ixi = ix*m;
|
||||
jyi = jy*m;
|
||||
iz1 = ixi+jyi;
|
||||
iv1 = jyi-ixi;
|
||||
// iz2 = (n*m-iz1);
|
||||
// iv2 = n*m-iv1;
|
||||
for (i=0;i<m;i++,ixi++,jyi++,iz1++,iv1++) {
|
||||
rrA = rA[ixi]*rA[jyi]; ///rrA = rA[i][ix]*rA[i][jy];
|
||||
iiA = iA[ixi]*iA[jyi]; ///iiA = iA[i][ix]*iA[i][jy];
|
||||
riA = rA[ixi]*iA[jyi]; ///riA = rA[i][ix]*iA[i][jy];
|
||||
irA = iA[ixi]*rA[jyi]; ///irA = iA[i][ix]*rA[i][jy];
|
||||
|
||||
/* Sum frequency effects */
|
||||
tmp1 = kfact*2.0*(rrA-iiA)*Epij;
|
||||
tmp2 = kfact*2.0*(riA+irA)*Epij;
|
||||
rvec[iz1] += tmp1;///rvec[i][jy+ix] += tmp1;
|
||||
ivec[iz1] += tmp2;///ivec[i][jy+ix] += tmp2;
|
||||
//rvec[iz2] += tmp1;///rvec[i][n*m-(jy+ix)] += tmp1;
|
||||
//ivec[iz2] -= tmp2;///ivec[i][n*m-(jy+ix)] -= tmp2;
|
||||
//iz2++;
|
||||
|
||||
/* Difference frequency effects */
|
||||
tmp1 = kfact*2.0*(rrA+iiA)*Edij;
|
||||
tmp2 = kfact*2.0*(riA-irA)*Edij;
|
||||
rvec[iv1] += tmp1;///rvec[i][jy-ix] += tmp1;
|
||||
ivec[iv1] += tmp2;///ivec[i][jy-ix] -= tmp2;
|
||||
|
||||
//rvec[iv2] += tmp1;
|
||||
//ivec[iv2] -= tmp2;
|
||||
//iv2++;
|
||||
}
|
||||
}
|
||||
}
|
||||
}
|
||||
//return i;
|
||||
}
|
||||
/*
|
||||
* DISUFQ2 Is an internal function to spec2nlsdat
|
||||
*
|
||||
* CALL: disufq2(rsvec,isvec,rdvec,idvec,rA,iA, w,kw,h,g,nmin,nmax,m,n)
|
||||
*
|
||||
* rsvec, isvec = real and imaginary parts of the sum frequency
|
||||
* effects (size m X n).
|
||||
* rdvec, idvec = real and imaginary parts of the difference frequency
|
||||
* effects (size m X n).
|
||||
* rA, iA = real and imaginary parts of the amplitudes (size m X n).
|
||||
* w = vector with angular frequencies (w>=0)
|
||||
* kw = vector with wavenumbers (kw>=0)
|
||||
* h = water depth (h >=0)
|
||||
* g = constant acceleration of gravity
|
||||
* nmin = minimum index where rA(:,nmin) and iA(:,nmin) is
|
||||
* greater than zero.
|
||||
* nmax = maximum index where rA(:,nmax) and iA(:,nmax) is
|
||||
* greater than zero.
|
||||
* m = size(rA,1),size(iA,1)
|
||||
* n = size(rA,2),size(iA,2), or size(rvec,2),size(ivec,2)
|
||||
*
|
||||
* DISUFQ2 returns the summation of sum and difference frequency
|
||||
* frequency effects in the vectors svec = rsvec +sqrt(-1)*isvec and
|
||||
* dvec = rdvec +sqrt(-1)*idvec.
|
||||
* The 2'nd order contribution to the Stokes wave is then calculated by
|
||||
* a simple 1D Fourier transform, real(FFT(svec+dvec)).
|
||||
*
|
||||
*
|
||||
* This is a MEX-file for MATLAB.
|
||||
* by Per Andreas Brodtkorb 15.08.2001
|
||||
* revised pab 14.03.2002, 01.05.2002
|
||||
*/
|
||||
|
||||
void disufq2(double *rsvec, double *isvec,
|
||||
double *rdvec, double *idvec,
|
||||
double *rA, double *iA,
|
||||
double *w, double *kw,
|
||||
double h, double g,
|
||||
int nmin, int nmax,
|
||||
int m, int n)
|
||||
{
|
||||
double Epij, Edij;
|
||||
double tmp1, tmp2, tmp3, tmp4, kfact;
|
||||
double w1, w2, kw1, kw2, Cg;
|
||||
double rrA, iiA, riA, irA;
|
||||
int i,jy,ix,iz1,iv1,ixi,jyi;
|
||||
//int iz2,iv2
|
||||
|
||||
//Initialize rvec and ivec to zero
|
||||
for (ix=0;ix<n*m;ix++) {
|
||||
rsvec[ix] = 0.0;
|
||||
isvec[ix] = 0.0;
|
||||
rdvec[ix] = 0.0;
|
||||
idvec[ix] = 0.0;
|
||||
}
|
||||
|
||||
// kfact is set to 2 in order to exploit the symmetry.
|
||||
// If you set kfact to 1, you must uncomment all statements
|
||||
// including the expressions: rvec[iz2], rvec[iv2], ivec[iz2] and ivec[iv2].
|
||||
|
||||
kfact = 2.0;
|
||||
if (h>10000){ /* deep water /Inifinite water depth */
|
||||
for (ix = nmin-1;ix<nmax;ix++) {
|
||||
ixi = ix*m;
|
||||
iz1 = 2*ixi;
|
||||
//iz2 = n*m-ixi;
|
||||
kw1 = kw[ix];
|
||||
Epij = kw1;
|
||||
for (i=0;i<m;i++,ixi++,iz1++) {
|
||||
rrA = rA[ixi]*rA[ixi]; ///
|
||||
iiA = iA[ixi]*iA[ixi]; ///
|
||||
riA = rA[ixi]*iA[ixi]; ///
|
||||
|
||||
/// Sum frequency effects along the diagonal
|
||||
tmp1 = kfact*(rrA-iiA)*Epij;
|
||||
tmp2 = kfact*2.0*riA*Epij;
|
||||
rsvec[iz1] += tmp1;
|
||||
isvec[iz1] += tmp2;
|
||||
|
||||
//rsvec[iz2] += tmp1;
|
||||
//isvec[iz2] -= tmp2;
|
||||
//iz2++;
|
||||
|
||||
/// Difference frequency effects are zero along the diagonal
|
||||
/// and are thus not contributing to the mean.
|
||||
}
|
||||
for (jy = ix+1;jy<nmax;jy++){
|
||||
kw2 = kw[jy];
|
||||
Epij = 0.5*(kw2 + kw1);
|
||||
Edij = -0.5*(kw2 - kw1);
|
||||
//printf("Edij = %f Epij = %f \n", Edij,Epij);
|
||||
|
||||
ixi = ix*m;
|
||||
jyi = jy*m;
|
||||
iz1 = ixi+jyi;
|
||||
iv1 = jyi-ixi;
|
||||
//iz2 = (n*m-iz1);
|
||||
//iv2 = (n*m-iv1);
|
||||
for (i = 0;i<m;i++,ixi++,jyi++,iz1++,iv1++) {
|
||||
|
||||
rrA = rA[ixi]*rA[jyi]; ///rrA = rA[i][ix]*rA[i][jy];
|
||||
iiA = iA[ixi]*iA[jyi]; ///iiA = iA[i][ix]*iA[i][jy];
|
||||
riA = rA[ixi]*iA[jyi]; ///riA = rA[i][ix]*iA[i][jy];
|
||||
irA = iA[ixi]*rA[jyi]; ///irA = iA[i][ix]*rA[i][jy];
|
||||
|
||||
/* Sum frequency effects */
|
||||
tmp1 = kfact*2.0*(rrA-iiA)*Epij;
|
||||
tmp2 = kfact*2.0*(riA+irA)*Epij;
|
||||
rsvec[iz1] += tmp1; ///rvec[i][ix+jy] += tmp1;
|
||||
isvec[iz1] += tmp2; ///ivec[i][ix+jy] += tmp2;
|
||||
//rsvec[iz2] += tmp1;///rvec[i][n*m-(ix+jy)] += tmp1;
|
||||
//isvec[iz2] -= tmp2;///ivec[i][n*m-(ix+jy)] += tmp2;
|
||||
//iz2++;
|
||||
|
||||
/* Difference frequency effects */
|
||||
tmp1 = kfact*2.0*(rrA+iiA)*Edij;
|
||||
tmp2 = kfact*2.0*(riA-irA)*Edij;
|
||||
|
||||
rdvec[iv1] += tmp1;///rvec[i][jy-ix] += tmp1;
|
||||
idvec[iv1] += tmp2;///ivec[i][jy-ix] += tmp2;
|
||||
|
||||
//rdvec[iv2] += tmp1;///rvec[i][n*m-(jy-ix)] += tmp1;
|
||||
//idvec[iv2] -= tmp2;///ivec[i][n*m-(jy-ix)] -= tmp2;
|
||||
// iv2++;
|
||||
}
|
||||
}
|
||||
}
|
||||
}
|
||||
else{ /* Finite water depth */
|
||||
for (ix = nmin-1;ix<nmax;ix++) {
|
||||
kw1 = kw[ix];
|
||||
w1 = w[ix];
|
||||
tmp1 = tanh(kw1*h);
|
||||
/// Cg, wave group velocity
|
||||
Cg = 0.5*g*(tmp1 + kw1*h*(1.0- tmp1*tmp1))/w1; /// OK
|
||||
tmp1 = 0.5*g*(kw1/w1)*(kw1/w1);
|
||||
tmp2 = 0.5*w1*w1/g;
|
||||
tmp3 = g*kw1/(w1*Cg);
|
||||
|
||||
if (kw1*h<300.0){
|
||||
tmp4 = kw1/sinh(2.0*kw1*h);
|
||||
}
|
||||
else{ // To ensure sinh does not overflow.
|
||||
tmp4 = 0.0;
|
||||
}
|
||||
// Difference frequency effects finite water depth
|
||||
Edij = (tmp1-tmp2+tmp3)/(1.0-g*h/(Cg*Cg))-tmp4; /// OK
|
||||
|
||||
// Sum frequency effects finite water depth
|
||||
Epij = (3.0*(tmp1-tmp2)/(1.0-tmp1/kw1*tanh(2.0*kw1*h))+3.0*tmp2-tmp1); /// OK
|
||||
//printf("Edij = %f Epij = %f \n", Edij,Epij);
|
||||
|
||||
ixi = ix*m;
|
||||
iz1 = 2*ixi;
|
||||
//iz2 = n*m-ixi;
|
||||
for (i=0;i<m;i++,ixi++,iz1++) {
|
||||
|
||||
rrA = rA[ixi]*rA[ixi]; ///
|
||||
iiA = iA[ixi]*iA[ixi]; ///
|
||||
riA = rA[ixi]*iA[ixi]; ///
|
||||
|
||||
|
||||
/// Sum frequency effects along the diagonal
|
||||
rsvec[iz1] += kfact*(rrA-iiA)*Epij;
|
||||
isvec[iz1] += kfact*2.0*riA*Epij;
|
||||
//rsvec[iz2] += kfact*(rrA-iiA)*Epij;
|
||||
//isvec[iz2] -= kfact*2.0*riA*Epij;
|
||||
|
||||
/// Difference frequency effects along the diagonal
|
||||
/// are only contributing to the mean
|
||||
//printf(" %f \n",2.0*(rrA+iiA)*Edij);
|
||||
rdvec[i] += 2.0*(rrA+iiA)*Edij;
|
||||
}
|
||||
for (jy = ix+1;jy<nmax;jy++) {
|
||||
// w1 = w[ix];
|
||||
// kw1 = kw[ix];
|
||||
w2 = w[jy];
|
||||
kw2 = kw[jy];
|
||||
tmp1 = g*(kw1/w1)*(kw2/w2);
|
||||
tmp2 = 0.5/g*(w1*w1+w2*w2+w1*w2);
|
||||
tmp3 = 0.5*g*(w1*kw2*kw2+w2*kw1*kw1)/(w1*w2*(w1+w2));
|
||||
tmp4 = (1-g*(kw1+kw2)/(w1+w2)/(w1+w2)*tanh((kw1+kw2)*h));
|
||||
Epij = (tmp1-tmp2+tmp3)/tmp4+tmp2-0.5*tmp1; /* OK */
|
||||
|
||||
tmp2 = 0.5/g*(w1*w1+w2*w2-w1*w2); /*OK*/
|
||||
tmp3 = -0.5*g*(w1*kw2*kw2-w2*kw1*kw1)/(w1*w2*(w1-w2));
|
||||
tmp4 = (1.0-g*(kw1-kw2)/(w1-w2)/(w1-w2)*tanh((kw1-kw2)*h));
|
||||
Edij = (tmp1-tmp2+tmp3)/tmp4+tmp2-0.5*tmp1; /* OK */
|
||||
//printf("Edij = %f Epij = %f \n", Edij,Epij);
|
||||
|
||||
ixi = ix*m;
|
||||
jyi = jy*m;
|
||||
iz1 = ixi+jyi;
|
||||
iv1 = jyi-ixi;
|
||||
// iz2 = (n*m-iz1);
|
||||
// iv2 = (n*m-iv1);
|
||||
for (i=0;i<m;i++,ixi++,jyi++,iz1++,iv1++) {
|
||||
rrA = rA[ixi]*rA[jyi]; ///rrA = rA[i][ix]*rA[i][jy];
|
||||
iiA = iA[ixi]*iA[jyi]; ///iiA = iA[i][ix]*iA[i][jy];
|
||||
riA = rA[ixi]*iA[jyi]; ///riA = rA[i][ix]*iA[i][jy];
|
||||
irA = iA[ixi]*rA[jyi]; ///irA = iA[i][ix]*rA[i][jy];
|
||||
|
||||
/* Sum frequency effects */
|
||||
tmp1 = kfact*2.0*(rrA-iiA)*Epij;
|
||||
tmp2 = kfact*2.0*(riA+irA)*Epij;
|
||||
rsvec[iz1] += tmp1;///rsvec[i][jy+ix] += tmp1;
|
||||
isvec[iz1] += tmp2;///isvec[i][jy+ix] += tmp2;
|
||||
//rsvec[iz2] += tmp1;///rsvec[i][n*m-(jy+ix)] += tmp1;
|
||||
//isvec[iz2] -= tmp2;///isvec[i][n*m-(jy-ix)] += tmp2;
|
||||
//iz2++;
|
||||
|
||||
/* Difference frequency effects */
|
||||
tmp1 = kfact*2.0*(rrA+iiA)*Edij;
|
||||
tmp2 = kfact*2.0*(riA-irA)*Edij;
|
||||
rdvec[iv1] += tmp1;
|
||||
idvec[iv1] += tmp2;
|
||||
|
||||
//rdvec[iv2] += tmp1;
|
||||
//idvec[iv2] -= tmp2;
|
||||
// iv2++;
|
||||
}
|
||||
}
|
||||
}
|
||||
}
|
||||
// return i;
|
||||
}
|
||||
|
||||
|
||||
/* ++++++++++ BEGIN RF3 [ampl ampl_mean nr_of_cycle] */
|
||||
/* ++++++++++ Rain flow without time analysis */
|
||||
//By Adam Nieslony
|
||||
//Visit the MATLAB Central File Exchange for latest version
|
||||
//http://www.mathworks.com/matlabcentral/fileexchange/3026
|
||||
void findrfc3_astm(double *array_ext, double *array_out, int n, int *nout) {
|
||||
|
||||
double *pr, *po, a[16384], ampl, mean;
|
||||
int tot_num, index, j, cNr1, cNr2;
|
||||
|
||||
tot_num = n;
|
||||
|
||||
// pointers to the first element of the arrays
|
||||
pr = &array_ext[0];
|
||||
po = &array_out[0];
|
||||
|
||||
// The original rainflow counting by Nieslony, unchanged
|
||||
j = -1;
|
||||
cNr1 = 1;
|
||||
for (index=0; index<tot_num; index++) {
|
||||
a[++j]=*pr++;
|
||||
while ( (j >= 2) && (fabs(a[j-1]-a[j-2]) <= fabs(a[j]-a[j-1])) ) {
|
||||
ampl=fabs( (a[j-1]-a[j-2])/2 );
|
||||
switch(j) {
|
||||
case 0: { break; }
|
||||
case 1: { break; }
|
||||
case 2: {
|
||||
mean=(a[0]+a[1])/2;
|
||||
a[0]=a[1];
|
||||
a[1]=a[2];
|
||||
j=1;
|
||||
if (ampl > 0) {
|
||||
*po++=ampl;
|
||||
*po++=mean;
|
||||
*po++=0.50;
|
||||
}
|
||||
break;
|
||||
}
|
||||
default: {
|
||||
mean=(a[j-1]+a[j-2])/2;
|
||||
a[j-2]=a[j];
|
||||
j=j-2;
|
||||
if (ampl > 0) {
|
||||
*po++=ampl;
|
||||
*po++=mean;
|
||||
*po++=1.00;
|
||||
cNr1++;
|
||||
}
|
||||
break;
|
||||
}
|
||||
}
|
||||
}
|
||||
}
|
||||
cNr2 = 1;
|
||||
for (index=0; index<j; index++) {
|
||||
ampl=fabs(a[index]-a[index+1])/2;
|
||||
mean=(a[index]+a[index+1])/2;
|
||||
if (ampl > 0){
|
||||
*po++=ampl;
|
||||
*po++=mean;
|
||||
*po++=0.50;
|
||||
cNr2++;
|
||||
}
|
||||
}
|
||||
// array of ints nout is outputted
|
||||
nout[0] = cNr1;
|
||||
nout[1] = cNr2;
|
||||
}
|
||||
/* ++++++++++ END RF3 */
|
||||
|
||||
|
||||
// ++ BEGIN RF5 [ampl ampl_mean nr_of_cycle cycle_begin_time cycle_period_time]
|
||||
/* ++++++++++ Rain flow with time analysis */
|
||||
//By Adam Nieslony
|
||||
//Visit the MATLAB Central File Exchange for latest version
|
||||
//http://www.mathworks.com/matlabcentral/fileexchange/3026
|
||||
void
|
||||
findrfc5_astm(double *array_ext, double *array_t, double *array_out, int n, int *nout) {
|
||||
double *pr, *pt, *po, a[16384], t[16384], ampl, mean, period, atime;
|
||||
int tot_num, index, j, cNr1, cNr2;
|
||||
|
||||
|
||||
// tot_num = mxGetM(array_ext) * mxGetN(array_ext);
|
||||
tot_num = n;
|
||||
|
||||
// pointers to the first element of the arrays
|
||||
pr = &array_ext[0];
|
||||
pt = &array_t[0];
|
||||
po = &array_out[0];
|
||||
|
||||
// array_out = mxCreateDoubleMatrix(5, tot_num-1, mxREAL);
|
||||
|
||||
// The original rainflow counting by Nieslony, unchanged
|
||||
j = -1;
|
||||
cNr1 = 1;
|
||||
for (index=0; index<tot_num; index++) {
|
||||
a[++j]=*pr++;
|
||||
t[j]=*pt++;
|
||||
while ( (j >= 2) && (fabs(a[j-1]-a[j-2]) <= fabs(a[j]-a[j-1])) ) {
|
||||
ampl=fabs( (a[j-1]-a[j-2])/2 );
|
||||
switch(j)
|
||||
{
|
||||
case 0: { break; }
|
||||
case 1: { break; }
|
||||
case 2: {
|
||||
mean=(a[0]+a[1])/2;
|
||||
period=(t[1]-t[0])*2;
|
||||
atime=t[0];
|
||||
a[0]=a[1];
|
||||
a[1]=a[2];
|
||||
t[0]=t[1];
|
||||
t[1]=t[2];
|
||||
j=1;
|
||||
if (ampl > 0) {
|
||||
*po++=ampl;
|
||||
*po++=mean;
|
||||
*po++=0.50;
|
||||
*po++=atime;
|
||||
*po++=period;
|
||||
}
|
||||
break;
|
||||
}
|
||||
default: {
|
||||
mean=(a[j-1]+a[j-2])/2;
|
||||
period=(t[j-1]-t[j-2])*2;
|
||||
atime=t[j-2];
|
||||
a[j-2]=a[j];
|
||||
t[j-2]=t[j];
|
||||
j=j-2;
|
||||
if (ampl > 0) {
|
||||
*po++=ampl;
|
||||
*po++=mean;
|
||||
*po++=1.00;
|
||||
*po++=atime;
|
||||
*po++=period;
|
||||
cNr1++;
|
||||
}
|
||||
break;
|
||||
}
|
||||
}
|
||||
}
|
||||
}
|
||||
cNr2 = 1;
|
||||
for (index=0; index<j; index++) {
|
||||
ampl=fabs(a[index]-a[index+1])/2;
|
||||
mean=(a[index]+a[index+1])/2;
|
||||
period=(t[index+1]-t[index])*2;
|
||||
atime=t[index];
|
||||
if (ampl > 0){
|
||||
*po++=ampl;
|
||||
*po++=mean;
|
||||
*po++=0.50;
|
||||
*po++=atime;
|
||||
*po++=period;
|
||||
cNr2++;
|
||||
}
|
||||
}
|
||||
// /* free the memeory !!!*/
|
||||
// mxSetN(array_out, tot_num - cNr);
|
||||
nout[0] = cNr1;
|
||||
nout[1] = cNr2;
|
||||
}
|
||||
/* ++++++++++ END RF5 */
|
@ -1,24 +0,0 @@
|
||||
Copyright (c) 2003, Adam Niesłony
|
||||
All rights reserved.
|
||||
|
||||
Redistribution and use in source and binary forms, with or without
|
||||
modification, are permitted provided that the following conditions are
|
||||
met:
|
||||
|
||||
* Redistributions of source code must retain the above copyright
|
||||
notice, this list of conditions and the following disclaimer.
|
||||
* Redistributions in binary form must reproduce the above copyright
|
||||
notice, this list of conditions and the following disclaimer in
|
||||
the documentation and/or other materials provided with the distribution
|
||||
|
||||
THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
|
||||
AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
|
||||
IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
|
||||
ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE
|
||||
LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
|
||||
CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
|
||||
SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
|
||||
INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
|
||||
CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
|
||||
ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
|
||||
POSSIBILITY OF SUCH DAMAGE.
|
@ -1,86 +0,0 @@
|
||||
! File c_code.pyf
|
||||
python module c_library
|
||||
interface
|
||||
subroutine findrfc(y1, hmin, ind, n, info)
|
||||
intent(c) findrfc ! findrfc is a C function
|
||||
intent(c) ! all findrfc arguments are considered as C based
|
||||
integer intent(hide), depend(y1) :: n=len(y1)
|
||||
double precision dimension(n), intent(in) :: y1 ! input array
|
||||
double precision intent(in) :: hmin
|
||||
integer dimension(n), intent(out) :: ind ! output array,
|
||||
integer dimension(1), intent(out) :: info
|
||||
end subroutine findrfc
|
||||
subroutine findcross(y, v, ind, n, info)
|
||||
intent(c) findcross ! findcross is a C function
|
||||
intent(c) ! all findcross arguments are considered as C based
|
||||
integer intent(hide), depend(y) :: n=len(y)
|
||||
double precision dimension(n), intent(in) :: y ! input array
|
||||
double precision intent(in) :: v
|
||||
integer dimension(n), intent(out) :: ind ! output array,
|
||||
integer dimension(1),intent(out) :: info
|
||||
end subroutine findcross
|
||||
subroutine disufq(rvec, ivec, rA, iA, w, kw, h, g,nmin,nmax, m, n)
|
||||
intent(c) disufq ! disufq is a C function
|
||||
intent(c) ! all disufq arguments are considered as C based
|
||||
!integer intent(hide), depend(rA),check(n*m==len(iA)) :: n=len(rA)/m
|
||||
!integer intent(hide), depend(rA), check(m==shape(iA,1)) :: m=shape(rA,1)
|
||||
double precision dimension(n*m), intent(in) :: rA, iA ! input array
|
||||
double precision dimension(n/2+1), intent(in) :: w, kw ! input array
|
||||
double precision intent(in) :: h, g
|
||||
integer intent(in) :: nmin, nmax
|
||||
double precision dimension(n*m), intent(out) :: rvec, ivec ! output array,
|
||||
end subroutine disufq
|
||||
subroutine disufq2(rsvec, isvec,rdvec, idvec, rA, iA, w, kw, h, g,nmin,nmax, m, n)
|
||||
intent(c) disufq2 ! disufq2 is a C function
|
||||
intent(c) ! all disufq2 arguments are considered as C based
|
||||
!integer intent(hide), depend(rA),check(n*m==len(iA)) :: n=len(rA)/m
|
||||
!integer intent(hide), depend(rA), check(m==shape(iA,1)) :: m=shape(rA,1)
|
||||
double precision dimension(n*m), intent(in) :: rA, iA ! input array
|
||||
double precision dimension(n/2+1), intent(in) :: w, kw ! input array
|
||||
double precision intent(in) :: h, g
|
||||
integer intent(in) :: nmin, nmax
|
||||
double precision dimension(n*m), intent(out) :: rsvec, isvec, rdvec, idvec ! output array,
|
||||
end subroutine disufq2
|
||||
|
||||
! ===== START NIESLONY RAINFLOW FUNCTIONS
|
||||
! RAINFLOW Revision: 1.1
|
||||
! by Adam Nieslony, 2009
|
||||
subroutine findrfc3_astm(array_ext, array_out, n, nout)
|
||||
intent(c) findrfc3_astm ! rf3 is a C function
|
||||
intent(c) ! all rf3 arguments are
|
||||
! considered as C based
|
||||
|
||||
! n is the length of the input array array_ext
|
||||
integer intent(hide), depend(array_ext) :: n=len(array_ext)
|
||||
|
||||
! of input array x
|
||||
double precision intent(in) :: array_ext(n)
|
||||
|
||||
! the output array
|
||||
double precision intent(out) :: array_out(n,3)
|
||||
|
||||
! nout array, to output additional ints
|
||||
integer dimension(2), intent(out) :: nout
|
||||
end subroutine findrfc3_astm
|
||||
|
||||
subroutine findrfc5_astm(array_ext, array_t, array_out, n, nout)
|
||||
intent(c) findrfc5_astm ! rf5 is a C function
|
||||
intent(c) ! all rf5 arguments are
|
||||
! considered as C based
|
||||
|
||||
! n is the length of the input array array_ext
|
||||
integer intent(hide), depend(array_ext) :: n=len(array_ext)
|
||||
|
||||
! of input array x
|
||||
double precision intent(in) :: array_ext(n), array_t(n)
|
||||
|
||||
! the output array
|
||||
double precision intent(out) :: array_out(n,5)
|
||||
|
||||
! nout array, to output additional ints
|
||||
integer dimension(2), intent(out) :: nout
|
||||
end subroutine findrfc5_astm
|
||||
! ===== END NIESLONY RAINFLOW FUNCTIONS
|
||||
|
||||
end interface
|
||||
end python module c_library
|
File diff suppressed because it is too large
Load Diff
@ -1,18 +0,0 @@
|
||||
'''
|
||||
python setup.py build_src build_ext --inplace
|
||||
|
||||
See also http://www.scipy.org/Cookbook/CompilingExtensionsOnWindowsWithMinGW
|
||||
'''
|
||||
# File setup.py
|
||||
|
||||
|
||||
def configuration(parent_package='', top_path=None):
|
||||
from numpy.distutils.misc_util import Configuration
|
||||
config = Configuration('', parent_package, top_path)
|
||||
|
||||
config.add_extension('c_library',
|
||||
sources=['c_library.pyf', 'c_functions.c'])
|
||||
return config
|
||||
if __name__ == "__main__":
|
||||
from numpy.distutils.core import setup
|
||||
setup(**configuration(top_path='').todict())
|
@ -1,450 +0,0 @@
|
||||
PROGRAM sp2Acdf1
|
||||
C***********************************************************************
|
||||
C This program computes upper and lower bounds for: *
|
||||
C *
|
||||
C density of T_i, for Ac <=h, in a gaussian process i.e. *
|
||||
C *
|
||||
C half wavelength (up-crossing to downcrossing) for crests <h *
|
||||
C or half wavelength (down-crossing to upcrossing) for trough >h *
|
||||
C I.R. 27 Dec. 1999 *
|
||||
C***********************************************************************
|
||||
use GLOBALDATA, only : Nt,Nj,Nd,Nc,Ntd,Ntdc,NI,Mb,
|
||||
&NIT,Nx,TWOPI,XSPLT,SCIS,NSIMmax,COV
|
||||
use rind
|
||||
IMPLICIT NONE
|
||||
double precision, dimension(:,:),allocatable :: BIG
|
||||
double precision, dimension(:,:),allocatable :: ansrup
|
||||
double precision, dimension(:,:),allocatable :: ansrlo
|
||||
double precision, dimension(: ),allocatable :: ex,CY
|
||||
double precision, dimension(:,:),allocatable :: xc,fxind
|
||||
double precision, dimension(: ),allocatable :: h
|
||||
double precision, dimension(: ),allocatable :: R0,R1,R2,R3,R4
|
||||
double precision ::CC,U,XddInf,XdInf,XtInf
|
||||
double precision, dimension(:,:),allocatable :: a_up,a_lo
|
||||
integer , dimension(: ),allocatable :: seed
|
||||
integer ,dimension(7) :: indI
|
||||
integer :: Nstart,Ntime,tn,ts,speed,ph,def,seed1,seed_size,icy
|
||||
integer ::it1,it2,status
|
||||
double precision :: ds,dT ! lag spacing for covariances
|
||||
! f90 sp2Acdf1.f rind50.f
|
||||
|
||||
CALL INIT_LEVELS(U,def,Ntime,Nstart,NIT,speed,Nx,dT)
|
||||
!print *,'U,def,Ntime,Nstart,NIT,speed,SCIS,seed1,Nx,dT'
|
||||
!print *,U,def,Ntime,Nstart,NIT,speed,SCIS,seed1,Nx,dT
|
||||
!XSPLT=1.5d0
|
||||
if (SCIS.GT.0) then
|
||||
allocate(COV(1:Nx))
|
||||
call random_seed(SIZE=seed_size)
|
||||
allocate(seed(seed_size))
|
||||
call random_seed(GET=seed(1:seed_size)) ! get current seed
|
||||
seed(1)=seed1 ! change seed
|
||||
call random_seed(PUT=seed(1:seed_size))
|
||||
deallocate(seed)
|
||||
endif
|
||||
CALL INITDATA(speed)
|
||||
!print *,ntime,speed,u,NIT
|
||||
allocate(R0(1:Ntime+1))
|
||||
allocate(R1(1:Ntime+1))
|
||||
allocate(R2(1:Ntime+1))
|
||||
if (abs(def).GT.1) THEN
|
||||
allocate(R3(1:Ntime+1))
|
||||
allocate(R4(1:Ntime+1))
|
||||
!CALL INIT_AMPLITUDES(h,def,Nx)
|
||||
endif
|
||||
allocate(h(1:Nx))
|
||||
CALL INIT_AMPLITUDES(h,def,Nx)
|
||||
CALL INIT_COVARIANCES(Ntime,def,R0,R1,R2,R3,R4)
|
||||
|
||||
NI=4; Nd=2
|
||||
Nc=3; Mb=2
|
||||
|
||||
Nj=0
|
||||
indI(1)=0
|
||||
C ***** The bound 'infinity' is set to 10*sigma *****
|
||||
XdInf=10.d0*SQRT(-R2(1))
|
||||
XtInf=10.d0*SQRT(R0(1))
|
||||
! normalizing constant
|
||||
CC=TWOPI*SQRT(-R0(1)/R2(1))*exp(u*u/(2.d0*R0(1)) )
|
||||
|
||||
allocate(CY(1:Nx))
|
||||
do icy=1,Nx
|
||||
CY(icy)=exp(-0.5*h(icy)*h(icy)/100)/(10*sqrt(twopi))
|
||||
enddo
|
||||
allocate(BIG(1:Ntime+Nc,1:Ntime+Nc),stat=status)
|
||||
if (status.ne.0) then
|
||||
print *,'can not allocate BIG'
|
||||
end if
|
||||
allocate(ex(1:Ntime+Nc),stat=status)
|
||||
if (status.ne.0) then
|
||||
print *,'can not allocate ex'
|
||||
end if
|
||||
allocate(ansrup(1:Ntime,1:Nx))
|
||||
allocate(ansrlo(1:Ntime,1:Nx))
|
||||
ansrup=0.d0
|
||||
ansrlo=0.d0
|
||||
allocate(fxind(1:Nx,1:2))
|
||||
fxind=0.d0 !this is not needed
|
||||
allocate(xc(1:Nc,1:Nx))
|
||||
|
||||
|
||||
allocate(a_up(Mb,NI-1))
|
||||
allocate(a_lo(Mb,NI-1))
|
||||
a_up=0.d0
|
||||
a_lo=0.d0
|
||||
xc(1,1:Nx)=h(1:Nx)
|
||||
xc(2,1:Nx)=u
|
||||
xc(3,1:Nx)=u
|
||||
|
||||
if (def.GT.0) then
|
||||
a_up(1,1)=0.d0
|
||||
a_lo(1,1)=u
|
||||
a_up(1,2)=XdInf
|
||||
a_lo(1,3)=-XdInf
|
||||
a_up(2,1)=1.d0
|
||||
else
|
||||
a_up(1,1)=u
|
||||
a_lo(1,1)=0.d0
|
||||
a_lo(1,2)=-XdInf
|
||||
a_up(1,3)= XdInf
|
||||
a_lo(2,1)=1.d0
|
||||
endif
|
||||
!print *,'Nstart',Nstart
|
||||
Nstart=MAX(3,Nstart)
|
||||
|
||||
|
||||
if (SCIS.GT.0) then
|
||||
open (unit=11, file='COV.out', STATUS='unknown')
|
||||
write(11,*) 0.d0
|
||||
endif
|
||||
|
||||
!print *,'loop starts'
|
||||
do Ntd=Nstart,Ntime
|
||||
|
||||
Ntdc=Ntd+Nc
|
||||
ex=0.d0
|
||||
BIG=0.d0
|
||||
CALL COV_INPUT(BIG(1:Ntdc,1:Ntdc),Ntd,-1,R0,R1,R2,R3,R4) ! positive wave period
|
||||
|
||||
Nt=Ntd-Nd;
|
||||
indI(2)=Nt;
|
||||
indI(3)=Nt+1;
|
||||
indI(4)=Ntd;
|
||||
|
||||
CALL RINDD(fxind,Big(1:Ntdc,1:Ntdc),ex,xc,indI,a_lo,a_up)
|
||||
!print *,'test',fxind/CY(1:Nx)
|
||||
|
||||
do icy=1,Nx
|
||||
ansrup(Ntd,icy)=fxind(icy,1)*CC/CY(icy)
|
||||
ansrlo(Ntd,icy)=fxind(icy,2)*CC/CY(icy)
|
||||
enddo
|
||||
if (SCIS.GT.0) then
|
||||
write(11,*) COV(1) ! save coefficient of variation
|
||||
endif
|
||||
if((Nx.gt.4).or.NIT.gt.4) print *,'Ready: ',Ntd,' of ',Ntime
|
||||
enddo
|
||||
goto 300
|
||||
300 open (unit=11, file='dens.out', STATUS='unknown')
|
||||
|
||||
do ts=1,Ntime
|
||||
do ph=1,Nx
|
||||
write(11,*) ansrup(ts,ph),ansrlo(ts,ph)
|
||||
enddo
|
||||
enddo
|
||||
!111 FORMAT(2x,F12.8)
|
||||
close(11)
|
||||
900 continue
|
||||
deallocate(BIG)
|
||||
deallocate(ex)
|
||||
deallocate(fxind)
|
||||
deallocate(ansrup)
|
||||
deallocate(ansrlo)
|
||||
deallocate(xc)
|
||||
deallocate(R0)
|
||||
deallocate(R1)
|
||||
deallocate(R2)
|
||||
if (allocated(COV) ) then
|
||||
deallocate(COV)
|
||||
endif
|
||||
|
||||
if (allocated(R3)) then
|
||||
deallocate(R3)
|
||||
deallocate(R4)
|
||||
deallocate(h)
|
||||
ENDIF
|
||||
stop
|
||||
!return
|
||||
|
||||
CONTAINS
|
||||
|
||||
|
||||
|
||||
SUBROUTINE INIT_LEVELS
|
||||
& (U,def,Ntime,Nstart,NIT,speed,Nx,dT)
|
||||
IMPLICIT NONE
|
||||
integer, intent(out):: def,Ntime,Nstart,NIT,speed,Nx
|
||||
double precision ,intent(out) :: U,dT
|
||||
|
||||
|
||||
OPEN(UNIT=14,FILE='reflev.in',STATUS= 'UNKNOWN')
|
||||
READ (14,*) U
|
||||
READ (14,*) def
|
||||
READ (14,*) Ntime
|
||||
READ (14,*) Nstart
|
||||
READ (14,*) NIT
|
||||
READ (14,*) speed
|
||||
READ (14,*) SCIS
|
||||
READ (14,*) seed1
|
||||
READ (14,*) Nx
|
||||
|
||||
if (abs(def).GT.1) then
|
||||
READ (14,*) dT
|
||||
if (Ntime.lt.3) then
|
||||
print *,'The number of wavelength points is too small, stop'
|
||||
stop
|
||||
end if
|
||||
else
|
||||
if (Ntime.lt.2) then
|
||||
print *,'The number of wavelength points is too small, stop'
|
||||
stop
|
||||
end if
|
||||
endif
|
||||
CLOSE(UNIT=14)
|
||||
|
||||
RETURN
|
||||
END SUBROUTINE INIT_LEVELS
|
||||
|
||||
C******************************************************
|
||||
SUBROUTINE INIT_AMPLITUDES(h,def,Nx)
|
||||
IMPLICIT NONE
|
||||
double precision, dimension(:), intent(out) :: h
|
||||
integer, intent(in) :: def
|
||||
integer, intent(in) :: Nx
|
||||
integer :: ix
|
||||
|
||||
|
||||
OPEN(UNIT=4,FILE='h.in',STATUS= 'UNKNOWN')
|
||||
|
||||
C
|
||||
C Reading in amplitudes
|
||||
C
|
||||
do ix=1,Nx
|
||||
READ (4,*) H(ix)
|
||||
enddo
|
||||
CLOSE(UNIT=4)
|
||||
!if (def.LT.0) THEN
|
||||
! H=-H
|
||||
!endif
|
||||
|
||||
RETURN
|
||||
END SUBROUTINE INIT_AMPLITUDES
|
||||
|
||||
C**************************************************
|
||||
|
||||
C***********************************************************************
|
||||
C***********************************************************************
|
||||
|
||||
SUBROUTINE INIT_COVARIANCES(Ntime,def,R0,R1,R2,R3,R4)
|
||||
IMPLICIT NONE
|
||||
double precision, dimension(:),intent(out) :: R0,R1,R2
|
||||
double precision, dimension(:),intent(out) :: R3,R4
|
||||
integer,intent(in) :: Ntime,def
|
||||
integer :: i
|
||||
open (unit=1, file='Cd0.in',STATUS='unknown')
|
||||
open (unit=2, file='Cd1.in',STATUS='unknown')
|
||||
open (unit=3, file='Cd2.in',STATUS='unknown')
|
||||
|
||||
do i=1,Ntime
|
||||
read(1,*) R0(i)
|
||||
read(2,*) R1(i)
|
||||
read(3,*) R2(i)
|
||||
enddo
|
||||
close(1)
|
||||
close(2)
|
||||
close(3)
|
||||
|
||||
if (abs(def).GT.1) then
|
||||
open (unit=4, file='Cd3.in',STATUS='unknown')
|
||||
open (unit=5, file='Cd4.in',STATUS='unknown')
|
||||
|
||||
do i=1,Ntime
|
||||
read(4,*) R3(i)
|
||||
read(5,*) R4(i)
|
||||
enddo
|
||||
|
||||
close(4)
|
||||
close(5)
|
||||
endif
|
||||
return
|
||||
END SUBROUTINE INIT_COVARIANCES
|
||||
|
||||
C***********************************************************************
|
||||
C***********************************************************************
|
||||
|
||||
C**********************************************************************
|
||||
|
||||
SUBROUTINE COV_INPUT(BIG,tn,ts, R0,R1,R2,R3,R4)
|
||||
IMPLICIT NONE
|
||||
double precision, dimension(:,:),intent(inout) :: BIG
|
||||
double precision, dimension(:),intent(in) :: R0,R1,R2
|
||||
double precision, dimension(:),intent(in) :: R3,R4
|
||||
integer ,intent(in) :: tn,ts
|
||||
integer :: i,j,shft,Ntd1,N !=Ntdc
|
||||
double precision :: tmp
|
||||
! the order of the variables in the covariance matrix
|
||||
! are organized as follows:
|
||||
! For ts>1:
|
||||
! X(t2)..X(ts),..X(tn-1) X''(ts) X'(t1) X'(tn) X(ts) X(t1) X(tn) X'(ts)
|
||||
! = [Xt Xd Xc]
|
||||
!
|
||||
! For ts<=1:
|
||||
! X(t2)..,..X(tn-1) X'(t1) X'(tn) Y X(t1) X(tn)
|
||||
! = [Xt Xd Xc]
|
||||
!Add Y Condition : Y=h
|
||||
|
||||
! where
|
||||
!
|
||||
! Xt= time points in the indicator function
|
||||
! Xd= derivatives
|
||||
! Xc=variables to condition on
|
||||
|
||||
if (ts.LE.1) THEN
|
||||
Ntd1=tn
|
||||
N=Ntd1+Nc;
|
||||
shft=0 ! def=1 want only crest period Tc
|
||||
else
|
||||
Ntd1=tn+1
|
||||
N=Ntd1+4
|
||||
shft=1 ! def=2 or 3 want Tc Ac or Tcf, Ac
|
||||
endif
|
||||
|
||||
do i=1,tn-2
|
||||
!cov(Xt)
|
||||
do j=i,tn-2
|
||||
BIG(i,j) = R0(j-i+1) ! cov(X(ti+1),X(tj+1))
|
||||
enddo
|
||||
!cov(Xt,Xc)
|
||||
BIG(i ,Ntd1+1+shft) = 0.d0 !cov(X(ti+1),Y)
|
||||
BIG(i ,Ntd1+2+shft) = R0(i+1) !cov(X(ti+1),X(t1))
|
||||
BIG(tn-1-i ,Ntd1+3+shft) = R0(i+1) !cov(X(t.. ),X(tn))
|
||||
!Cov(Xt,Xd)=cov(X(ti+1),x(tj)
|
||||
BIG(i,Ntd1-1) =-R1(i+1) !cov(X(ti+1),X' (t1))
|
||||
BIG(tn-1-i,Ntd1)= R1(i+1) !cov(X(ti+1),X' (tn))
|
||||
enddo
|
||||
!call echo(big(1:tn,1:tn),tn)
|
||||
!cov(Xd)
|
||||
BIG(Ntd1 ,Ntd1 ) = -R2(1)
|
||||
BIG(Ntd1-1,Ntd1 ) = -R2(tn) !cov(X'(t1),X'(tn))
|
||||
BIG(Ntd1-1,Ntd1-1) = -R2(1)
|
||||
|
||||
!cov(Xc)
|
||||
!print *,'t'
|
||||
BIG(Ntd1+1+shft,Ntd1+1+shft) = 100.d0!100.d0 ! cov(Y,Y)
|
||||
BIG(Ntd1+1+shft,Ntd1+2+shft) = 0.d0
|
||||
BIG(Ntd1+1+shft,Ntd1+3+shft) = 0.d0
|
||||
BIG(Ntd1+2+shft,Ntd1+2+shft) = R0(1) ! cov(X(t1),X (t1))
|
||||
BIG(Ntd1+2+shft,Ntd1+3+shft) = R0(tn) ! cov(X(t1),X (tn))
|
||||
BIG(Ntd1+3+shft,Ntd1+3+shft) = R0(1) ! cov(X(tn),X (tn))
|
||||
!cov(Xd,Xc)
|
||||
BIG(Ntd1 ,Ntd1+1+shft) = 0.d0 !cov(X'(tn),Y)
|
||||
BIG(Ntd1 ,Ntd1+2+shft) = R1(tn) !cov(X'(tn),X(t1))
|
||||
BIG(Ntd1 ,Ntd1+3+shft) = 0.d0 !cov(X'(tn),X(tn))
|
||||
BIG(Ntd1-1,Ntd1+1+shft) = 0.d0 !cov(X'(t1),Y)
|
||||
BIG(Ntd1-1,Ntd1+2+shft) = 0.d0 !cov(X'(t1),X(t1))
|
||||
BIG(Ntd1-1,Ntd1+3+shft) =-R1(tn) !cov(X'(t1),X(tn))
|
||||
|
||||
|
||||
!call echo(big(1:N,1:N),N)
|
||||
! make lower triangular part equal to upper
|
||||
do j=1,N-1
|
||||
do i=j+1,N
|
||||
tmp =BIG(j,i)
|
||||
BIG(i,j)=tmp
|
||||
enddo
|
||||
!call echo(big(1:N,1:N),N)
|
||||
enddo
|
||||
!if (tn.eq.3) then
|
||||
!do j=1,N
|
||||
! do i=j,N
|
||||
! print *,'test',j,i,BIG(j,i)
|
||||
! enddo
|
||||
!call echo(big(1:N,1:N),N)
|
||||
!enddo
|
||||
!endif
|
||||
!call echo(big(1:N,1:N),N)
|
||||
|
||||
C write (*,10) ((BIG(j,i),i=N+1,N+6),j=N+1,N+6)
|
||||
C 10 format(6F8.4)
|
||||
RETURN
|
||||
END SUBROUTINE COV_INPUT
|
||||
|
||||
SUBROUTINE COV_INPUT2(BIG,pt, R0,R1,R2)
|
||||
IMPLICIT NONE
|
||||
double precision, dimension(:,:), intent(out) :: BIG
|
||||
double precision, dimension(:), intent(in) :: R0,R1,R2
|
||||
integer :: pt,i,j
|
||||
! the order of the variables in the covariance matrix
|
||||
! are organized as follows;
|
||||
! X(t2)...X(tn-1) X'(t1) X'(tn) X(t1) X(tn) = [Xt Xd Xc]
|
||||
!
|
||||
! where Xd is the derivatives
|
||||
!
|
||||
! Xt= time points in the indicator function
|
||||
! Xd= derivatives
|
||||
! Xc=variables to condition on
|
||||
|
||||
!cov(Xc)
|
||||
BIG(pt+2,pt+2) = R0(1)
|
||||
BIG(pt+1,pt+1) = R0(1)
|
||||
BIG(pt+1,pt+2) = R0(pt)
|
||||
!cov(Xd)
|
||||
BIG(pt,pt) = -R2(1)
|
||||
BIG(pt-1,pt-1) = -R2(1)
|
||||
BIG(pt-1,pt) = -R2(pt)
|
||||
!cov(Xd,Xc)
|
||||
BIG(pt,pt+2) = 0.d0
|
||||
BIG(pt,pt+1) = R1(pt)
|
||||
BIG(pt-1,pt+2) = -R1(pt)
|
||||
BIG(pt-1,pt+1) = 0.d0
|
||||
|
||||
if (pt.GT.2) then
|
||||
!cov(Xt)
|
||||
do i=1,pt-2
|
||||
do j=i,pt-2
|
||||
BIG(i,j) = R0(j-i+1)
|
||||
enddo
|
||||
enddo
|
||||
!cov(Xt,Xc)
|
||||
do i=1,pt-2
|
||||
BIG(i,pt+1) = R0(i+1)
|
||||
BIG(pt-1-i,pt+2) = R0(i+1)
|
||||
enddo
|
||||
!Cov(Xt,Xd)=cov(X(ti+1),x(tj))
|
||||
do i=1,pt-2
|
||||
BIG(i,pt-1) = -R1(i+1)
|
||||
BIG(pt-1-i,pt)= R1(i+1)
|
||||
enddo
|
||||
endif
|
||||
|
||||
|
||||
! make lower triangular part equal to upper
|
||||
do j=1,pt+1
|
||||
do i=j+1,pt+2
|
||||
BIG(i,j)=BIG(j,i)
|
||||
enddo
|
||||
enddo
|
||||
C write (*,10) ((BIG(j,i),i=N+1,N+6),j=N+1,N+6)
|
||||
C 10 format(6F8.4)
|
||||
RETURN
|
||||
END SUBROUTINE COV_INPUT2
|
||||
|
||||
|
||||
END PROGRAM sp2Acdf1
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
@ -1,356 +0,0 @@
|
||||
PROGRAM sp2mM1
|
||||
C***********************************************************************************
|
||||
C Computes upper lower bounds for density of maximum and the following minimum *
|
||||
C***********************************************************************************
|
||||
use GLOBALDATA, only : Nt,Nj,Nd,Nc,Ntd,Ntdc,NI,Mb,
|
||||
&NIT,Nx,TWOPI,XSPLT,SCIS,NSIMmax,COV
|
||||
use rind
|
||||
IMPLICIT NONE
|
||||
double precision, dimension(:,:),allocatable :: BIG
|
||||
double precision, dimension(:,:),allocatable :: ansrup
|
||||
double precision, dimension(:,:),allocatable :: ansrlo
|
||||
double precision, dimension(: ),allocatable :: ex,h
|
||||
double precision, dimension(:,:),allocatable :: xc
|
||||
double precision, dimension(:,:),allocatable :: fxind
|
||||
double precision, dimension(: ),allocatable :: R0,R1,R2,R3,R4
|
||||
double precision ::CC,U,XddInf,XdInf,XtInf
|
||||
double precision, dimension(:,:),allocatable :: a_up,a_lo
|
||||
integer , dimension(: ),allocatable :: seed
|
||||
integer ,dimension(7) :: indI
|
||||
integer :: Nstart,Ntime,tn,ts,speed,seed1,seed_size
|
||||
integer :: status,i,j,ij,Nx1
|
||||
double precision :: ds,dT ! lag spacing for covariances
|
||||
|
||||
! f90 sp2AmM1.f rind52.f
|
||||
|
||||
CALL INIT_LEVELS(Ntime,Nstart,NIT,speed,Nx1,dT)
|
||||
Nx=Nx1*(Nx1-1)/2
|
||||
if (SCIS.GT.0) then
|
||||
allocate(COV(1:Nx))
|
||||
call random_seed(SIZE=seed_size)
|
||||
allocate(seed(seed_size))
|
||||
call random_seed(GET=seed(1:seed_size)) ! get current seed
|
||||
seed(1)=seed1 ! change seed
|
||||
call random_seed(PUT=seed(1:seed_size))
|
||||
deallocate(seed)
|
||||
endif
|
||||
|
||||
CALL INITDATA(speed)
|
||||
|
||||
allocate(R0(1:Ntime+1))
|
||||
allocate(R1(1:Ntime+1))
|
||||
allocate(R2(1:Ntime+1))
|
||||
allocate(R3(1:Ntime+1))
|
||||
allocate(R4(1:Ntime+1))
|
||||
allocate(h(1:Nx1))
|
||||
|
||||
CALL INIT_AMPLITUDES(h,Nx1)
|
||||
CALL INIT_COVARIANCES(Ntime,R0,R1,R2,R3,R4)
|
||||
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
|
||||
! Y= X'(t2)...X'(tn-1)||X''(t1) X''(tn)||X(t1) X(tn) X'(t1) X'(tn) !!
|
||||
! = [ Xt Xd Xc ] !!
|
||||
! !!
|
||||
! Nt=tn-2, Nd=2, Nc=4 !!
|
||||
! !!
|
||||
! Xt= contains Nt time points in the indicator function !!
|
||||
! Xd= " Nd derivatives !!
|
||||
! Xc= " Nc variables to condition on !!
|
||||
! !!
|
||||
! There are 3 ( NI=4) regions with constant bariers: !!
|
||||
! (indI(1)=0); for i\in (indI(1),indI(2)] Y(i)<0. !!
|
||||
! (indI(2)=Nt) ; for i\in (indI(2)+1,indI(3)], Y(i)<0 (deriv. X''(t1)) !!
|
||||
! (indI(3)=Nt+1); for i\in (indI(3)+1,indI(4)], Y(i)>0 (deriv. X''(tn)) !!
|
||||
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
|
||||
|
||||
|
||||
NI=4; Nd=2
|
||||
Nc=4; Mb=1
|
||||
|
||||
Nj=0
|
||||
indI(1)=0
|
||||
C ***** The bound 'infinity' is set to 10*sigma *****
|
||||
XdInf=10.d0*SQRT(R4(1))
|
||||
XtInf=10.d0*SQRT(-R2(1))
|
||||
! normalizing constant
|
||||
CC=TWOPI*SQRT(-R2(1)/R4(1))
|
||||
|
||||
allocate(BIG(1:Ntime+Nc,1:Ntime+Nc),stat=status)
|
||||
if (status.ne.0) then
|
||||
print *,'can not allocate BIG'
|
||||
end if
|
||||
allocate(ex(1:Ntime+Nc),stat=status)
|
||||
if (status.ne.0) then
|
||||
print *,'can not allocate ex'
|
||||
end if
|
||||
allocate(ansrup(1:Nx1,1:Nx1))
|
||||
ansrup=0.d0
|
||||
allocate(ansrlo(1:Nx1,1:Nx1))
|
||||
ansrlo=0.d0
|
||||
allocate(fxind(1:Nx,1:2))
|
||||
fxind=0.d0 !this is not needed
|
||||
allocate(xc(1:Nc,1:Nx))
|
||||
|
||||
|
||||
allocate(a_up(Mb,NI-1))
|
||||
allocate(a_lo(Mb,NI-1))
|
||||
|
||||
a_up=0.d0
|
||||
a_lo=0.d0
|
||||
|
||||
ij=0
|
||||
do i=2,Nx1
|
||||
do j=1,i-1
|
||||
ij=ij+1
|
||||
xc(1,ij)=h(i)
|
||||
xc(2,ij)=h(j)
|
||||
enddo
|
||||
enddo
|
||||
xc(3,1:Nx)=0.d0
|
||||
xc(4,1:Nx)=0.d0
|
||||
|
||||
a_lo(1,1)=-Xtinf
|
||||
a_lo(1,2)=-XdInf
|
||||
a_up(1,3)=+XdInf
|
||||
|
||||
|
||||
Nstart=MAX(2,Nstart)
|
||||
|
||||
|
||||
if (SCIS.GT.0) then
|
||||
open (unit=11, file='COV.out', STATUS='unknown')
|
||||
write(11,*) 0.d0
|
||||
endif
|
||||
|
||||
do Ntd=Nstart,Ntime
|
||||
|
||||
Ntdc=Ntd+Nc
|
||||
ex=0.d0
|
||||
BIG=0.d0
|
||||
CALL COV_INPUT(BIG(1:Ntdc,1:Ntdc),Ntd,R0,R1,R2,R3,R4) ! positive wave period
|
||||
|
||||
Nt=Ntd-Nd;
|
||||
indI(2)=Nt;
|
||||
indI(3)=Nt+1;
|
||||
indI(4)=Ntd;
|
||||
|
||||
CALL RINDD(fxind,Big(1:Ntdc,1:Ntdc),ex,xc,indI,a_lo,a_up)
|
||||
ij=0
|
||||
do i=2,Nx1
|
||||
do j=1,i-1
|
||||
ij=ij+1
|
||||
ansrup(i,j)=ansrup(i,j)+fxind(ij,1)*CC*dt
|
||||
ansrlo(i,j)=ansrlo(i,j)+fxind(ij,2)*CC*dt
|
||||
enddo
|
||||
enddo
|
||||
|
||||
if (SCIS.GT.0) then
|
||||
write(11,*) COV(1) ! save coefficient of variation
|
||||
endif
|
||||
print *,'Ready: ',Ntd,' of ',Ntime
|
||||
enddo
|
||||
goto 300
|
||||
300 open (unit=11, file='dens.out', STATUS='unknown')
|
||||
do i=1,Nx1
|
||||
do j=1,Nx1
|
||||
write(11,*) ansrup(i,j),ansrlo(i,j)
|
||||
enddo
|
||||
enddo
|
||||
close(11)
|
||||
900 continue
|
||||
deallocate(BIG)
|
||||
deallocate(ex)
|
||||
deallocate(fxind)
|
||||
deallocate(ansrup)
|
||||
deallocate(ansrlo)
|
||||
deallocate(xc)
|
||||
deallocate(R0)
|
||||
deallocate(R1)
|
||||
deallocate(R2)
|
||||
deallocate(R3)
|
||||
deallocate(R4)
|
||||
deallocate(h)
|
||||
|
||||
if (allocated(COV) ) then
|
||||
deallocate(COV)
|
||||
endif
|
||||
stop
|
||||
!return
|
||||
|
||||
CONTAINS
|
||||
|
||||
|
||||
|
||||
SUBROUTINE INIT_LEVELS
|
||||
& (Ntime,Nstart,NIT,speed,Nx,dT)
|
||||
IMPLICIT NONE
|
||||
integer, intent(out):: Ntime,Nstart,NIT,speed,Nx
|
||||
double precision ,intent(out) :: dT
|
||||
|
||||
|
||||
OPEN(UNIT=14,FILE='reflev.in',STATUS= 'UNKNOWN')
|
||||
READ (14,*) Ntime
|
||||
READ (14,*) Nstart
|
||||
READ (14,*) NIT
|
||||
READ (14,*) speed
|
||||
READ (14,*) SCIS
|
||||
READ (14,*) seed1
|
||||
READ (14,*) Nx
|
||||
READ (14,*) dT
|
||||
|
||||
if (Ntime.lt.2) then
|
||||
print *,'The number of wavelength points is too small, stop'
|
||||
stop
|
||||
end if
|
||||
CLOSE(UNIT=14)
|
||||
|
||||
RETURN
|
||||
END SUBROUTINE INIT_LEVELS
|
||||
|
||||
C******************************************************
|
||||
SUBROUTINE INIT_AMPLITUDES(h,Nx)
|
||||
IMPLICIT NONE
|
||||
double precision, dimension(:), intent(out) :: h
|
||||
integer, intent(in) :: Nx
|
||||
integer :: ix
|
||||
|
||||
|
||||
OPEN(UNIT=4,FILE='h.in',STATUS= 'UNKNOWN')
|
||||
|
||||
C
|
||||
C Reading in amplitudes
|
||||
C
|
||||
do ix=1,Nx
|
||||
READ (4,*) H(ix)
|
||||
enddo
|
||||
CLOSE(UNIT=4)
|
||||
|
||||
RETURN
|
||||
END SUBROUTINE INIT_AMPLITUDES
|
||||
|
||||
C**************************************************
|
||||
|
||||
C***********************************************************************
|
||||
C***********************************************************************
|
||||
|
||||
SUBROUTINE INIT_COVARIANCES(Ntime,R0,R1,R2,R3,R4)
|
||||
IMPLICIT NONE
|
||||
double precision, dimension(:),intent(out) :: R0,R1,R2
|
||||
double precision, dimension(:),intent(out) :: R3,R4
|
||||
integer,intent(in) :: Ntime
|
||||
integer :: i
|
||||
open (unit=1, file='Cd0.in',STATUS='unknown')
|
||||
open (unit=2, file='Cd1.in',STATUS='unknown')
|
||||
open (unit=3, file='Cd2.in',STATUS='unknown')
|
||||
open (unit=4, file='Cd3.in',STATUS='unknown')
|
||||
open (unit=5, file='Cd4.in',STATUS='unknown')
|
||||
|
||||
do i=1,Ntime
|
||||
read(1,*) R0(i)
|
||||
read(2,*) R1(i)
|
||||
read(3,*) R2(i)
|
||||
read(4,*) R3(i)
|
||||
read(5,*) R4(i)
|
||||
enddo
|
||||
close(1)
|
||||
close(2)
|
||||
close(3)
|
||||
close(3)
|
||||
close(5)
|
||||
return
|
||||
END SUBROUTINE INIT_COVARIANCES
|
||||
|
||||
C**********************************************************************
|
||||
|
||||
SUBROUTINE COV_INPUT(BIG,tn,R0,R1,R2,R3,R4)
|
||||
IMPLICIT NONE
|
||||
double precision, dimension(:,:),intent(inout) :: BIG
|
||||
double precision, dimension(:),intent(in) :: R0,R1,R2
|
||||
double precision, dimension(:),intent(in) :: R3,R4
|
||||
integer ,intent(in) :: tn
|
||||
integer :: i,j,N
|
||||
double precision :: tmp
|
||||
! the order of the variables in the covariance matrix
|
||||
! are organized as follows:
|
||||
! X'(t2)..X'(ts),...,X'(tn-1) X''(t1),X''(tn) X(t1),X(tn),X'(t1),X'(tn)
|
||||
! = [ Xt | Xd | Xc ]
|
||||
!
|
||||
! where
|
||||
!
|
||||
! Xt= time points in the indicator function
|
||||
! Xd= derivatives
|
||||
! Xc=variables to condition on
|
||||
|
||||
! Computations of all covariances follows simple rules: Cov(X(t),X(s))=r(t,s),
|
||||
! then Cov(X'(t),X(s))=dr(t,s)/dt. Now for stationary X(t) we have
|
||||
! a function r(tau) such that Cov(X(t),X(s))=r(s-t) (or r(t-s) will give the same result).
|
||||
!
|
||||
! Consequently Cov(X'(t),X(s)) = -r'(s-t) = -sign(s-t)*r'(|s-t|)
|
||||
! Cov(X'(t),X'(s)) = -r''(s-t) = -r''(|s-t|)
|
||||
! Cov(X''(t),X'(s)) = r'''(s-t) = sign(s-t)*r'''(|s-t|)
|
||||
! Cov(X''(t),X(s)) = r''(s-t) = r''(|s-t|)
|
||||
! Cov(X''(t),X''(s)) = r''''(s-t) = r''''(|s-t|)
|
||||
|
||||
N=tn+4
|
||||
do i=1,tn-2
|
||||
!cov(Xt)
|
||||
do j=i,tn-2
|
||||
BIG(i,j) = -R2(j-i+1) ! cov(X'(ti+1),X'(tj+1))
|
||||
enddo
|
||||
!cov(Xt,Xc)
|
||||
BIG(i ,tn+1) = R1(i+1) !cov(X'(ti+1),X(t1))
|
||||
BIG(tn-1-i ,tn+2) = -R1(i+1) !cov(X'(ti+1),X(tn))
|
||||
BIG(i ,tn+3) = -R2(i+1) !cov(X'(ti+1),X'(t1))
|
||||
BIG(tn-1-i ,tn+4) = -R2(i+1) !cov(X'(ti+1),X'(tn))
|
||||
!Cov(Xt,Xd)
|
||||
BIG(i,tn-1) = R3(i+1) !cov(X'(ti+1),X''(t1))
|
||||
BIG(tn-1-i,tn) =-R3(i+1) !cov(X'(ti+1),X''(tn))
|
||||
enddo
|
||||
|
||||
!cov(Xd)
|
||||
BIG(tn-1 ,tn-1 ) = R4(1)
|
||||
BIG(tn-1,tn ) = R4(tn) !cov(X''(t1),X''(tn))
|
||||
BIG(tn ,tn ) = R4(1)
|
||||
|
||||
!cov(Xc)
|
||||
BIG(tn+1,tn+1) = R0(1) ! cov(X(t1),X(t1))
|
||||
BIG(tn+1,tn+2) = R0(tn) ! cov(X(t1),X(tn))
|
||||
BIG(tn+1,tn+3) = 0.d0 ! cov(X(t1),X'(t1))
|
||||
BIG(tn+1,tn+4) = R1(tn) ! cov(X(t1),X'(tn))
|
||||
BIG(tn+2,tn+2) = R0(1) ! cov(X(tn),X(tn))
|
||||
BIG(tn+2,tn+3) =-R1(tn) ! cov(X(tn),X'(t1))
|
||||
BIG(tn+2,tn+4) = 0.d0 ! cov(X(tn),X'(tn))
|
||||
BIG(tn+3,tn+3) =-R2(1) ! cov(X'(t1),X'(t1))
|
||||
BIG(tn+3,tn+4) =-R2(tn) ! cov(X'(t1),X'(tn))
|
||||
BIG(tn+4,tn+4) =-R2(1) ! cov(X'(tn),X'(tn))
|
||||
!Xc=X(t1),X(tn),X'(t1),X'(tn)
|
||||
!Xd=X''(t1),X''(tn)
|
||||
!cov(Xd,Xc)
|
||||
BIG(tn-1 ,tn+1) = R2(1) !cov(X''(t1),X(t1))
|
||||
BIG(tn-1 ,tn+2) = R2(tn) !cov(X''(t1),X(tn))
|
||||
BIG(tn-1 ,tn+3) = 0.d0 !cov(X''(t1),X'(t1))
|
||||
BIG(tn-1 ,tn+4) = R3(tn) !cov(X''(t1),X'(tn))
|
||||
BIG(tn ,tn+1) = R2(tn) !cov(X''(tn),X(t1))
|
||||
BIG(tn ,tn+2) = R2(1) !cov(X''(tn),X(tn))
|
||||
BIG(tn ,tn+3) =-R3(tn) !cov(X''(tn),X'(t1))
|
||||
BIG(tn ,tn+4) = 0.d0 !cov(X''(tn),X'(tn))
|
||||
! make lower triangular part equal to upper
|
||||
do j=1,N-1
|
||||
do i=j+1,N
|
||||
tmp =BIG(j,i)
|
||||
BIG(i,j)=tmp
|
||||
enddo
|
||||
enddo
|
||||
RETURN
|
||||
END SUBROUTINE COV_INPUT
|
||||
|
||||
|
||||
END PROGRAM sp2mM1
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
@ -1,504 +0,0 @@
|
||||
PROGRAM sp2tccpdf1
|
||||
C***********************************************************************
|
||||
C This program computes upper and lower bounds for the: *
|
||||
C *
|
||||
C density of T= T_1+T_2 in a gaussian process i.e. *
|
||||
C *
|
||||
C wavelengthes for crests <h1 and troughs >h2 *
|
||||
C *
|
||||
C Sylvie and Igor 7 dec. 1999 *
|
||||
C***********************************************************************
|
||||
use GLOBALDATA, only : Nt,Nj,Nd,Nc,Ntd,Ntdc,NI,Mb,
|
||||
& NIT,Nx,TWOPI,XSPLT,SCIS,NSIMmax,COV
|
||||
use rind
|
||||
IMPLICIT NONE
|
||||
double precision, dimension(:,:),allocatable :: BIG
|
||||
double precision, dimension(:,:),allocatable :: ansrup
|
||||
double precision, dimension(:,:),allocatable :: ansrlo
|
||||
double precision, dimension(: ),allocatable :: ex,CY1,CY2
|
||||
double precision, dimension(:,:),allocatable :: xc
|
||||
double precision, dimension(:,:),allocatable ::fxind
|
||||
double precision, dimension(: ),allocatable :: h1,h2
|
||||
double precision, dimension(: ),allocatable :: hh1,hh2
|
||||
double precision, dimension(: ),allocatable :: R0,R1,R2
|
||||
double precision ::CC,U,XddInf,XdInf,XtInf
|
||||
double precision, dimension(:,:),allocatable :: a_up,a_lo
|
||||
integer , dimension(: ),allocatable :: seed
|
||||
integer ,dimension(7) :: indI
|
||||
integer :: Ntime,N0,tn,ts,speed,ph,seed1,seed_size,Nx1,Nx2
|
||||
integer :: icy,icy2
|
||||
double precision :: ds,dT ! lag spacing for covariances
|
||||
! DIGITAL:
|
||||
! f90 -g2 -C -automatic -o ~/WAT/V4/sp2tthpdf.exe rind48.f sp2tthpdf.f
|
||||
! SOLARIS:
|
||||
!f90 -g -O -w3 -Bdynamic -fixed -o ../sp2tthpdf.exe rind48.f sp2tthpdf.f
|
||||
|
||||
!print *,'enter sp2thpdf'
|
||||
CALL INIT_LEVELS(U,Ntime,N0,NIT,speed,SCIS,seed1,Nx1,Nx2,dT)
|
||||
|
||||
!print *,'U,Ntime,NIT,speed,SCIS,seed1,Nx,dT'
|
||||
!print *,U,Ntime,NIT,speed,SCIS,seed1,Nx,dT
|
||||
!Nx1=1
|
||||
!Nx2=1
|
||||
|
||||
Nx=Nx1*Nx2
|
||||
!print *,'NN',Nx1,Nx2,Nx
|
||||
|
||||
|
||||
!XSPLT=1.5d0
|
||||
if (SCIS.GT.0) then
|
||||
allocate(COV(1:Nx))
|
||||
call random_seed(SIZE=seed_size)
|
||||
allocate(seed(seed_size))
|
||||
call random_seed(GET=seed(1:seed_size)) ! get current seed
|
||||
seed(1)=seed1 ! change seed
|
||||
call random_seed(PUT=seed(1:seed_size))
|
||||
deallocate(seed)
|
||||
endif
|
||||
CALL INITDATA(speed)
|
||||
!print *,ntime,speed,u,NIT
|
||||
allocate(R0(1:Ntime+1))
|
||||
allocate(R1(1:Ntime+1))
|
||||
allocate(R2(1:Ntime+1))
|
||||
|
||||
allocate(h1(1:Nx1))
|
||||
allocate(h2(1:Nx2))
|
||||
CALL INIT_AMPLITUDES(h1,Nx1,h2,Nx2)
|
||||
CALL INIT_COVARIANCES(Ntime,R0,R1,R2)
|
||||
|
||||
|
||||
allocate(hh1(1:Nx))
|
||||
allocate(hh2(1:Nx))
|
||||
!h transformation
|
||||
do icy=1,Nx1
|
||||
do icy2=1,Nx2
|
||||
hh1((icy-1)*Nx2+icy2)=h1(icy);
|
||||
hh2((icy-1)*Nx2+icy2)=h2(icy2);
|
||||
enddo
|
||||
enddo
|
||||
|
||||
Nj=0
|
||||
indI(1)=0
|
||||
|
||||
C ***** The bound 'infinity' is set to 10*sigma *****
|
||||
XdInf=10.d0*SQRT(-R2(1))
|
||||
XtInf=10.d0*SQRT(R0(1))
|
||||
!h1(1)=XtInf
|
||||
!h2(1)=XtInf
|
||||
! normalizing constant
|
||||
CC=TWOPI*SQRT(-R0(1)/R2(1))*exp(u*u/(2.d0*R0(1)) )
|
||||
allocate(CY1(1:Nx))
|
||||
allocate(CY2(1:Nx))
|
||||
do icy=1,Nx
|
||||
CY1(icy)=exp(-0.5*hh1(icy)*hh1(icy)/100)/(10*sqrt(twopi))
|
||||
CY2(icy)=exp(-0.5*hh2(icy)*hh2(icy)/100)/(10*sqrt(twopi))
|
||||
enddo
|
||||
!print *,CY1
|
||||
allocate(ansrup(1:Ntime,1:Nx))
|
||||
allocate(ansrlo(1:Ntime,1:Nx))
|
||||
ansrup=0.d0
|
||||
ansrlo=0.d0
|
||||
allocate(fxind(1:Nx,1:2))
|
||||
!fxind=0.d0 this is not needed
|
||||
|
||||
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
|
||||
! Y={X(t2)..,X(ts),..X(tn-1)||X'(ts) X'(t1) X'(tn)||Y1 Y2 X(ts) X(t1) X(tn)} !!
|
||||
! = [Xt Xd Xc] !!
|
||||
! !!
|
||||
! Nt=tn-2, Nd=3, Nc=2+3 !!
|
||||
! !!
|
||||
! Xt= contains Nt time points in the indicator function !!
|
||||
! Xd= " Nd derivatives !!
|
||||
! Xc= " Nc variables to condition on !!
|
||||
! (Y1,Y2) dummy variables ind. of all other v. inputing h1,h2 into rindd !!
|
||||
! !!
|
||||
! There are 6 ( NI=7) regions with constant bariers: !!
|
||||
! (indI(1)=0); for i\in (indI(1),indI(2)] u<Y(i)<h1 !!
|
||||
! (indI(2)=ts-2); for i\in (indI(2),indI(2)], inf<Y(i)<inf (no restr.) !!
|
||||
! (indI(3)=ts-1); for i\in (indI(3),indI(4)], h2 <Y(i)<u !!
|
||||
! (indI(4)=Nt) ; for i\in (indI(4),indI(5)], Y(i)<0 (deriv. X'(ts)) !!
|
||||
! (indI(5)=Nt+1); for i\in (indI(5),indI(6)], Y(i)>0 (deriv. X'(t1)) !!
|
||||
! (indI(6)=Nt+2); for i\in (indI(6),indI(7)], Y(i)>0 (deriv. X'(tn)) !!
|
||||
! (indI(7)=Nt+3); NI=7. !!
|
||||
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
|
||||
|
||||
|
||||
NI=7; Nd=3
|
||||
Nc=5; Mb=3
|
||||
allocate(a_up(1:Mb,1:(NI-1)))
|
||||
allocate(a_lo(1:Mb,1:(NI-1)))
|
||||
a_up=0.d0
|
||||
a_lo=0.d0
|
||||
allocate(BIG(1:(Ntime+Nc+1),1:(Ntime+Nc+1)))
|
||||
ALLOCATE(xc(1:Nc,1:Nx))
|
||||
allocate(ex(1:(Ntime+Nc+1)))
|
||||
!print *,size(ex),Ntime
|
||||
ex=0.d0
|
||||
!print *,size(ex),ex
|
||||
xc(1,1:Nx)=hh1(1:Nx)
|
||||
xc(2,1:Nx)=hh2(1:Nx)
|
||||
xc(3,1:Nx)=u
|
||||
xc(4,1:Nx)=u
|
||||
xc(5,1:Nx)=u
|
||||
! upp- down- upp-crossings at t1,ts,tn
|
||||
|
||||
a_lo(1,1)=u
|
||||
a_up(1,2)=XtInf ! X(ts) is redundant
|
||||
a_lo(1,2)=-Xtinf
|
||||
a_up(1,3)=u
|
||||
|
||||
|
||||
a_lo(1,4)=-XdInf
|
||||
a_up(1,5)= XdInf
|
||||
a_up(1,6)= XdInf
|
||||
|
||||
a_up(2,1)=1.d0
|
||||
a_lo(3,3)=1.d0 !signe a voir!!!!!!
|
||||
! print *,a_up
|
||||
! print *,a_lo
|
||||
do tn=N0,Ntime,1
|
||||
! do tn=Ntime,Ntime,1
|
||||
Ntd=tn+1
|
||||
Nt=Ntd-Nd
|
||||
Ntdc=Ntd+Nc
|
||||
indI(4)=Nt
|
||||
indI(5)=Nt+1
|
||||
indI(6)=Nt+2
|
||||
indI(7)=Ntd
|
||||
if (SCIS.gt.0) then
|
||||
if (SCIS.EQ.2) then
|
||||
Nj=max(Nt,0)
|
||||
else
|
||||
Nj=min(max(Nt-5, 0),0)
|
||||
endif
|
||||
endif
|
||||
do ts=3,tn-2
|
||||
!print *,'ts,tn' ,ts,tn,Ntdc
|
||||
CALL COV_INPUT(Big(1:Ntdc,1:Ntdc),tn,ts,R0,R1,R2)!positive wave period
|
||||
indI(2)=ts-2
|
||||
indI(3)=ts-1
|
||||
|
||||
|
||||
CALL RINDD(fxind,Big(1:Ntdc,1:Ntdc),ex(1:Ntdc),
|
||||
& xc,indI,a_lo,a_up)
|
||||
|
||||
ds=dt
|
||||
do icy=1,Nx
|
||||
! ansr(tn,:)=ansr(tn,:)+fxind*CC*ds./(CY1.*CY2)
|
||||
ansrup(tn,icy)=ansrup(tn,icy)+fxind(icy,1)*CC*ds
|
||||
& /(CY1(icy)*CY2(icy))
|
||||
ansrlo(tn,icy)=ansrlo(tn,icy)+fxind(icy,2)*CC*ds
|
||||
& /(CY1(icy)*CY2(icy))
|
||||
enddo
|
||||
enddo ! ts
|
||||
print *,'Ready: ',tn,' of ',Ntime
|
||||
|
||||
enddo !tn
|
||||
|
||||
300 open (unit=11, file='dens.out', STATUS='unknown')
|
||||
|
||||
do ts=1,Ntime
|
||||
do ph=1,Nx
|
||||
!write(11,*) ansrup(ts,ph),ansrlo(ts,ph)
|
||||
write(11,111) ansrup(ts,ph),ansrlo(ts,ph)
|
||||
enddo
|
||||
enddo
|
||||
111 FORMAT(2x,F12.8,2x,F12.8)
|
||||
close(11)
|
||||
900 deallocate(big)
|
||||
deallocate(fxind)
|
||||
deallocate(ansrup)
|
||||
deallocate(ansrlo)
|
||||
deallocate(xc)
|
||||
deallocate(ex)
|
||||
deallocate(R0)
|
||||
deallocate(R1)
|
||||
deallocate(R2)
|
||||
if (allocated(COV) ) then
|
||||
deallocate(COV)
|
||||
endif
|
||||
deallocate(h1)
|
||||
deallocate(h2)
|
||||
deallocate(hh1)
|
||||
deallocate(hh2)
|
||||
deallocate(a_up)
|
||||
deallocate(a_lo)
|
||||
stop
|
||||
!return
|
||||
|
||||
CONTAINS
|
||||
|
||||
|
||||
|
||||
SUBROUTINE INIT_LEVELS
|
||||
& (U,Ntime,N0,NIT,speed,SCIS,seed1,Nx1,Nx2,dT)
|
||||
IMPLICIT NONE
|
||||
integer, intent(out):: Ntime,N0,NIT,speed,Nx1,Nx2,SCIS,seed1
|
||||
double precision ,intent(out) :: U,dT
|
||||
|
||||
|
||||
OPEN(UNIT=14,FILE='reflev.in',STATUS= 'UNKNOWN')
|
||||
READ (14,*) U
|
||||
READ (14,*) Ntime
|
||||
READ (14,*) N0
|
||||
READ (14,*) NIT
|
||||
READ (14,*) speed
|
||||
READ (14,*) SCIS
|
||||
READ (14,*) seed1
|
||||
|
||||
|
||||
READ (14,*) Nx1,Nx2
|
||||
READ (14,*) dT
|
||||
if (Ntime.lt.5) then
|
||||
print *,'The number of wavelength points is too small, stop'
|
||||
stop
|
||||
end if
|
||||
|
||||
CLOSE(UNIT=14)
|
||||
|
||||
RETURN
|
||||
END SUBROUTINE INIT_LEVELS
|
||||
|
||||
C******************************************************
|
||||
SUBROUTINE INIT_AMPLITUDES(h1,Nx1,h2,Nx2)
|
||||
IMPLICIT NONE
|
||||
double precision, dimension(:), intent(out) :: h1,h2
|
||||
integer, intent(in) :: Nx1,Nx2
|
||||
integer :: ix
|
||||
|
||||
|
||||
OPEN(UNIT=4,FILE='h.in',STATUS= 'UNKNOWN')
|
||||
|
||||
C
|
||||
C Reading in amplitudes
|
||||
C
|
||||
do ix=1,Nx1
|
||||
READ (4,*) H1(ix)
|
||||
enddo
|
||||
do ix=1,Nx2
|
||||
READ (4,*) H2(ix)
|
||||
enddo
|
||||
CLOSE(UNIT=4)
|
||||
|
||||
RETURN
|
||||
END SUBROUTINE INIT_AMPLITUDES
|
||||
|
||||
C**************************************************
|
||||
|
||||
C***********************************************************************
|
||||
C***********************************************************************
|
||||
|
||||
SUBROUTINE INIT_COVARIANCES(Ntime,R0,R1,R2)
|
||||
IMPLICIT NONE
|
||||
double precision, dimension(:),intent(out) :: R0,R1,R2
|
||||
integer,intent(in) :: Ntime
|
||||
integer :: i
|
||||
open (unit=1, file='Cd0.in',STATUS='unknown')
|
||||
open (unit=2, file='Cd1.in',STATUS='unknown')
|
||||
open (unit=3, file='Cd2.in',STATUS='unknown')
|
||||
|
||||
do i=1,Ntime
|
||||
read(1,*) R0(i)
|
||||
read(2,*) R1(i)
|
||||
read(3,*) R2(i)
|
||||
enddo
|
||||
close(1)
|
||||
close(2)
|
||||
close(3)
|
||||
|
||||
return
|
||||
END SUBROUTINE INIT_COVARIANCES
|
||||
|
||||
C***********************************************************************
|
||||
C***********************************************************************
|
||||
|
||||
C**********************************************************************
|
||||
|
||||
SUBROUTINE COV_INPUT(BIG,tn,ts, R0,R1,R2)
|
||||
IMPLICIT NONE
|
||||
double precision, dimension(:,:),intent(inout) :: BIG
|
||||
double precision, dimension(:),intent(in) :: R0,R1,R2
|
||||
integer ,intent(in) :: tn,ts
|
||||
integer :: i,j,Ntd1,N !=Ntdc
|
||||
double precision :: tmp
|
||||
! the order of the variables in the covariance matrix
|
||||
! are organized as follows:
|
||||
!
|
||||
! ||X(t2)..X(ts),..X(tn-1)||X'(ts) X'(t1) X'(tn)||Y1 Y2 X(ts) X(t1) X(tn)||
|
||||
! = [Xt Xd Xc]
|
||||
! where
|
||||
!
|
||||
! Xt= time points in the indicator function
|
||||
! Xd= derivatives
|
||||
! Xc=variables to condition on
|
||||
|
||||
! Computations of all covariances follows simple rules: Cov(X(t),X(s))=r(t,s),
|
||||
! then Cov(X'(t),X(s))=dr(t,s)/dt. Now for stationary X(t) we have
|
||||
! a function r(tau) such that Cov(X(t),X(s))=r(s-t) (or r(t-s) will give the same result).
|
||||
!
|
||||
! Consequently Cov(X'(t),X(s)) = -r'(s-t) = -sign(s-t)*r'(|s-t|)
|
||||
! Cov(X'(t),X'(s)) = -r''(s-t) = -r''(|s-t|)
|
||||
! Cov(X''(t),X'(s)) = r'''(s-t) = sign(s-t)*r'''(|s-t|)
|
||||
! Cov(X''(t),X(s)) = r''(s-t) = r''(|s-t|)
|
||||
! Cov(X''(t),X''(s)) = r''''(s-t) = r''''(|s-t|)
|
||||
|
||||
Ntd1=tn+1
|
||||
N=Ntd1+Nc
|
||||
do i=1,tn-2
|
||||
!cov(Xt)
|
||||
do j=i,tn-2
|
||||
BIG(i,j) = R0(j-i+1) ! cov(X(ti+1),X(tj+1))
|
||||
enddo
|
||||
!cov(Xt,Xc)
|
||||
BIG(i ,Ntd1+1) = 0.d0 !cov(X(ti+1),Y1)
|
||||
BIG(i ,Ntd1+2) = 0.d0 !cov(X(ti+1),Y2)
|
||||
BIG(i ,Ntd1+4) = R0(i+1) !cov(X(ti+1),X(t1))
|
||||
BIG(tn-1-i ,Ntd1+5) = R0(i+1) !cov(X(t.. ),X(tn))
|
||||
|
||||
!Cov(Xt,Xd)=cov(X(ti+1),x(tj)
|
||||
BIG(i,Ntd1-1) =-R1(i+1) !cov(X(ti+1),X'(t1))
|
||||
BIG(tn-1-i,Ntd1)= R1(i+1) !cov(X(ti+1),X'(tn))
|
||||
enddo
|
||||
!cov(Xd)
|
||||
BIG(Ntd1 ,Ntd1 ) = -R2(1)
|
||||
BIG(Ntd1-1,Ntd1 ) = -R2(tn) !cov(X'(t1),X'(tn))
|
||||
BIG(Ntd1-1,Ntd1-1) = -R2(1)
|
||||
BIG(Ntd1-2,Ntd1-1) = -R2(ts) !cov(X'(ts),X'(t1))
|
||||
BIG(Ntd1-2,Ntd1-2) = -R2(1)
|
||||
BIG(Ntd1-2,Ntd1 ) = -R2(tn+1-ts) !cov(X'(ts),X'(tn))
|
||||
|
||||
!cov(Xc)
|
||||
BIG(Ntd1+1,Ntd1+1) = 100.d0 ! cov(Y1 Y1)
|
||||
BIG(Ntd1+1,Ntd1+2) = 0.d0 ! cov(Y1 Y2)
|
||||
BIG(Ntd1+1,Ntd1+3) = 0.d0 ! cov(Y1 X(ts))
|
||||
BIG(Ntd1+1,Ntd1+4) = 0.d0 ! cov(Y1 X(t1))
|
||||
BIG(Ntd1+1,Ntd1+5) = 0.d0 ! cov(Y1 X(tn))
|
||||
BIG(Ntd1+2,Ntd1+2) = 100.d0 ! cov(Y2 Y2)
|
||||
BIG(Ntd1+2,Ntd1+3) = 0.d0 ! cov(Y2 X(ts))
|
||||
BIG(Ntd1+2,Ntd1+4) = 0.d0 ! cov(Y2 X(t1))
|
||||
BIG(Ntd1+2,Ntd1+5) = 0.d0 ! cov(Y2 X(tn))
|
||||
|
||||
BIG(Ntd1+3,Ntd1+3) = R0(1) ! cov(X(ts),X (ts)
|
||||
BIG(Ntd1+3,Ntd1+4) = R0(ts) ! cov(X(ts),X (t1))
|
||||
BIG(Ntd1+3,Ntd1+5) = R0(tn+1-ts) ! cov(X(ts),X (tn))
|
||||
BIG(Ntd1+4,Ntd1+4) = R0(1) ! cov(X(t1),X (t1))
|
||||
BIG(Ntd1+4,Ntd1+5) = R0(tn) ! cov(X(t1),X (tn))
|
||||
BIG(Ntd1+5,Ntd1+5) = R0(1) ! cov(X(tn),X (tn))
|
||||
|
||||
|
||||
!cov(Xd,Xc)
|
||||
BIG(Ntd1 ,Ntd1+1) = 0.d0 !cov(X'(tn),Y1)
|
||||
BIG(Ntd1 ,Ntd1+2) = 0.d0 !cov(X'(tn),Y2)
|
||||
BIG(Ntd1-1 ,Ntd1+1) = 0.d0 !cov(X'(t1),Y1)
|
||||
BIG(Ntd1-1 ,Ntd1+2) = 0.d0 !cov(X'(t1),Y2)
|
||||
BIG(Ntd1-2 ,Ntd1+1) = 0.d0 !cov(X'(ts),Y1)
|
||||
BIG(Ntd1-2 ,Ntd1+2) = 0.d0 !cov(X'(ts),Y2)
|
||||
|
||||
BIG(Ntd1 ,Ntd1+4) = R1(tn) !cov(X'(tn),X(t1))
|
||||
BIG(Ntd1 ,Ntd1+5) = 0.d0 !cov(X'(tn),X(tn))
|
||||
BIG(Ntd1-1,Ntd1+4) = 0.d0 !cov(X'(t1),X(t1))
|
||||
BIG(Ntd1-1,Ntd1+5) =-R1(tn) !cov(X'(t1),X(tn))
|
||||
BIG(Ntd1 ,Ntd1+3) = R1(tn+1-ts) !cov(X'(tn),X (ts))
|
||||
BIG(Ntd1-1,Ntd1+3) =-R1(ts) !cov(X'(t1),X (ts))
|
||||
BIG(Ntd1-2,Ntd1+3) = 0.d0 !cov(X'(ts),X (ts)
|
||||
BIG(Ntd1-2,Ntd1+4) = R1(ts) !cov(X'(ts),X (t1))
|
||||
BIG(Ntd1-2,Ntd1+5) = -R1(tn+1-ts) !cov(X'(ts),X (tn))
|
||||
|
||||
|
||||
do i=1,tn-2
|
||||
j=abs(i+1-ts)
|
||||
!cov(Xt,Xc)
|
||||
BIG(i,Ntd1+3) = R0(j+1) !cov(X(ti+1),X(ts))
|
||||
!Cov(Xt,Xd)
|
||||
if ((i+1-ts).lt.0) then
|
||||
BIG(i,Ntd1-2) = R1(j+1)
|
||||
else !cov(X(ti+1),X'(ts))
|
||||
BIG(i,Ntd1-2) = -R1(j+1)
|
||||
endif
|
||||
enddo
|
||||
|
||||
! make lower triangular part equal to upper
|
||||
do j=1,N-1
|
||||
do i=j+1,N
|
||||
tmp =BIG(j,i)
|
||||
|
||||
BIG(i,j)=tmp
|
||||
enddo
|
||||
enddo
|
||||
|
||||
C write (*,10) ((BIG(j,i),i=N+1,N+6),j=N+1,N+6)
|
||||
C 10 format(6F8.4)
|
||||
RETURN
|
||||
END SUBROUTINE COV_INPUT
|
||||
|
||||
SUBROUTINE COV_INPUT2(BIG,pt, R0,R1,R2)
|
||||
IMPLICIT NONE
|
||||
double precision, dimension(:,:), intent(out) :: BIG
|
||||
double precision, dimension(:), intent(in) :: R0,R1,R2
|
||||
integer :: pt,i,j
|
||||
! the order of the variables in the covariance matrix
|
||||
! are organized as follows;
|
||||
! X(t2)...X(tn-1) X'(t1) X'(tn) X(t1) X(tn) = [Xt Xd Xc]
|
||||
!
|
||||
! where Xd is the derivatives
|
||||
!
|
||||
! Xt= time points in the indicator function
|
||||
! Xd= derivatives
|
||||
! Xc=variables to condition on
|
||||
|
||||
!cov(Xc)
|
||||
BIG(pt+2,pt+2) = R0(1)
|
||||
BIG(pt+1,pt+1) = R0(1)
|
||||
BIG(pt+1,pt+2) = R0(pt)
|
||||
!cov(Xd)
|
||||
BIG(pt,pt) = -R2(1)
|
||||
BIG(pt-1,pt-1) = -R2(1)
|
||||
BIG(pt-1,pt) = -R2(pt)
|
||||
!cov(Xd,Xc)
|
||||
BIG(pt,pt+2) = 0.d0
|
||||
BIG(pt,pt+1) = R1(pt)
|
||||
BIG(pt-1,pt+2) = -R1(pt)
|
||||
BIG(pt-1,pt+1) = 0.d0
|
||||
|
||||
if (pt.GT.2) then
|
||||
!cov(Xt)
|
||||
do i=1,pt-2
|
||||
do j=i,pt-2
|
||||
BIG(i,j) = R0(j-i+1)
|
||||
enddo
|
||||
enddo
|
||||
!cov(Xt,Xc)
|
||||
do i=1,pt-2
|
||||
BIG(i,pt+1) = R0(i+1)
|
||||
BIG(pt-1-i,pt+2) = R0(i+1)
|
||||
enddo
|
||||
!Cov(Xt,Xd)=cov(X(ti+1),x(tj))
|
||||
do i=1,pt-2
|
||||
BIG(i,pt-1) = -R1(i+1)
|
||||
BIG(pt-1-i,pt)= R1(i+1)
|
||||
enddo
|
||||
endif
|
||||
|
||||
|
||||
! make lower triangular part equal to upper
|
||||
do j=1,pt+1
|
||||
do i=j+1,pt+2
|
||||
BIG(i,j)=BIG(j,i)
|
||||
enddo
|
||||
enddo
|
||||
C write (*,10) ((BIG(j,i),i=N+1,N+6),j=N+1,N+6)
|
||||
C 10 format(6F8.4)
|
||||
RETURN
|
||||
END SUBROUTINE COV_INPUT2
|
||||
|
||||
|
||||
END PROGRAM sp2tccpdf1
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
@ -1,497 +0,0 @@
|
||||
PROGRAM sp2tthpdf1
|
||||
C***********************************************************************
|
||||
C This program computes: *
|
||||
C *
|
||||
C density of T= T_1+T_2 in a gaussian process i.e. *
|
||||
C *
|
||||
C wavelengthes for crests <h1 and troughs >h2 *
|
||||
C *
|
||||
C Sylvie and Igor 7 dec. 1999 *
|
||||
C***********************************************************************
|
||||
use GLOBALDATA, only : Nt,Nj,Nd,Nc,Ntd,Ntdc,NI,Mb,
|
||||
& NIT,Nx,TWOPI,XSPLT,SCIS,NSIMmax,COV
|
||||
use rind
|
||||
IMPLICIT NONE
|
||||
double precision, dimension(:,:),allocatable :: BIG
|
||||
double precision, dimension(:,:),allocatable :: ansr
|
||||
double precision, dimension(: ),allocatable :: ex,CY1,CY2
|
||||
double precision, dimension(:,:),allocatable :: xc
|
||||
double precision, dimension(: ),allocatable :: fxind,h1,h2
|
||||
double precision, dimension(: ),allocatable :: hh1,hh2
|
||||
double precision, dimension(: ),allocatable :: R0,R1,R2
|
||||
double precision ::CC,U,XddInf,XdInf,XtInf
|
||||
double precision, dimension(:,:),allocatable :: a_up,a_lo
|
||||
integer , dimension(: ),allocatable :: seed
|
||||
integer ,dimension(7) :: indI
|
||||
integer :: Ntime,tn,ts,speed,ph,seed1,seed_size,Nx1,Nx2,N0
|
||||
integer :: icy,icy2
|
||||
double precision :: ds,dT ! lag spacing for covariances
|
||||
! DIGITAL:
|
||||
! f90 -g2 -C -automatic -o ~/WAT/V4/sp2tthpdf1.exe rind49.f sp2tthpdf1.f
|
||||
! SOLARIS:
|
||||
!f90 -g -O -w3 -Bdynamic -fixed -o ../sp2tthpdf.exe rind49.f sp2tthpdf1.f
|
||||
|
||||
!print *,'enter sp2thpdf'
|
||||
CALL INIT_LEVELS(U,Ntime,N0,NIT,speed,SCIS,seed1,Nx1,Nx2,dT)
|
||||
|
||||
!print *,'U,Ntime,NIT,speed,SCIS,seed1,Nx,dT'
|
||||
!print *,U,Ntime,NIT,speed,SCIS,seed1,Nx,dT
|
||||
!Nx1=1
|
||||
!Nx2=1
|
||||
|
||||
Nx=Nx1*Nx2
|
||||
!print *,'NN',Nx1,Nx2,Nx
|
||||
|
||||
|
||||
!XSPLT=1.5d0
|
||||
if (SCIS.GT.0) then
|
||||
allocate(COV(1:Nx))
|
||||
call random_seed(SIZE=seed_size)
|
||||
allocate(seed(seed_size))
|
||||
call random_seed(GET=seed(1:seed_size)) ! get current seed
|
||||
seed(1)=seed1 ! change seed
|
||||
call random_seed(PUT=seed(1:seed_size))
|
||||
deallocate(seed)
|
||||
endif
|
||||
CALL INITDATA(speed)
|
||||
!print *,ntime,speed,u,NIT
|
||||
allocate(R0(1:Ntime+1))
|
||||
allocate(R1(1:Ntime+1))
|
||||
allocate(R2(1:Ntime+1))
|
||||
|
||||
allocate(h1(1:Nx1))
|
||||
allocate(h2(1:Nx2))
|
||||
CALL INIT_AMPLITUDES(h1,Nx1,h2,Nx2)
|
||||
CALL INIT_COVARIANCES(Ntime,R0,R1,R2)
|
||||
|
||||
|
||||
allocate(hh1(1:Nx))
|
||||
allocate(hh2(1:Nx))
|
||||
!h transformation
|
||||
do icy=1,Nx1
|
||||
do icy2=1,Nx2
|
||||
hh1((icy-1)*Nx2+icy2)=h1(icy);
|
||||
hh2((icy-1)*Nx2+icy2)=h2(icy2);
|
||||
enddo
|
||||
enddo
|
||||
|
||||
Nj=0
|
||||
indI(1)=0
|
||||
|
||||
C ***** The bound 'infinity' is set to 10*sigma *****
|
||||
XdInf=10.d0*SQRT(-R2(1))
|
||||
XtInf=10.d0*SQRT(R0(1))
|
||||
!h1(1)=XtInf
|
||||
!h2(1)=XtInf
|
||||
! normalizing constant
|
||||
CC=TWOPI*SQRT(-R0(1)/R2(1))*exp(u*u/(2.d0*R0(1)) )
|
||||
allocate(CY1(1:Nx))
|
||||
allocate(CY2(1:Nx))
|
||||
do icy=1,Nx
|
||||
CY1(icy)=exp(-0.5*hh1(icy)*hh1(icy)/100)/(10*sqrt(twopi))
|
||||
CY2(icy)=exp(-0.5*hh2(icy)*hh2(icy)/100)/(10*sqrt(twopi))
|
||||
enddo
|
||||
!print *,CY1
|
||||
allocate(ansr(1:Ntime,1:Nx))
|
||||
ansr=0.d0
|
||||
allocate(fxind(1:Nx))
|
||||
fxind=0.d0
|
||||
|
||||
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
|
||||
! Y={X(t2)..,X(ts),..X(tn-1)||X'(ts) X'(t1) X'(tn)||Y1 Y2 X(ts) X(t1) X(tn)} !!
|
||||
! = [Xt Xd Xc] !!
|
||||
! !!
|
||||
! Nt=tn-2, Nd=3, Nc=2+3 !!
|
||||
! !!
|
||||
! Xt= contains Nt time points in the indicator function !!
|
||||
! Xd= " Nd derivatives !!
|
||||
! Xc= " Nc variables to condition on !!
|
||||
! (Y1,Y2) dummy variables ind. of all other v. inputing h1,h2 into rindd !!
|
||||
! !!
|
||||
! There are 6 ( NI=7) regions with constant bariers: !!
|
||||
! (indI(1)=0); for i\in (indI(1),indI(2)] u<Y(i)<h1 !!
|
||||
! (indI(2)=ts-2); for i\in (indI(2),indI(2)], inf<Y(i)<inf (no restr.) !!
|
||||
! (indI(3)=ts-1); for i\in (indI(3),indI(4)], h2 <Y(i)<u !!
|
||||
! (indI(4)=Nt) ; for i\in (indI(4),indI(5)], Y(i)<0 (deriv. X'(ts)) !!
|
||||
! (indI(5)=Nt+1); for i\in (indI(5),indI(6)], Y(i)>0 (deriv. X'(t1)) !!
|
||||
! (indI(6)=Nt+2); for i\in (indI(6),indI(7)], Y(i)>0 (deriv. X'(tn)) !!
|
||||
! (indI(7)=Nt+3); NI=7. !!
|
||||
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
|
||||
|
||||
|
||||
NI=7; Nd=3
|
||||
Nc=5; Mb=3
|
||||
allocate(a_up(1:Mb,1:(NI-1)))
|
||||
allocate(a_lo(1:Mb,1:(NI-1)))
|
||||
a_up=0.d0
|
||||
a_lo=0.d0
|
||||
allocate(BIG(1:(Ntime+Nc+1),1:(Ntime+Nc+1)))
|
||||
ALLOCATE(xc(1:Nc,1:Nx))
|
||||
allocate(ex(1:(Ntime+Nc+1)))
|
||||
!print *,size(ex),Ntime
|
||||
ex=0.d0
|
||||
!print *,size(ex),ex
|
||||
xc(1,1:Nx)=hh1(1:Nx)
|
||||
xc(2,1:Nx)=hh2(1:Nx)
|
||||
xc(3,1:Nx)=u
|
||||
xc(4,1:Nx)=u
|
||||
xc(5,1:Nx)=u
|
||||
! upp- down- upp-crossings at t1,ts,tn
|
||||
|
||||
a_lo(1,1)=u
|
||||
a_up(1,2)=XtInf ! X(ts) is redundant
|
||||
a_lo(1,2)=-Xtinf
|
||||
a_up(1,3)=u
|
||||
|
||||
|
||||
a_lo(1,4)=-XdInf
|
||||
a_up(1,5)= XdInf
|
||||
a_up(1,6)= XdInf
|
||||
|
||||
a_up(2,1)=1.d0
|
||||
a_lo(3,3)=1.d0 !signe a voir!!!!!!
|
||||
! print *,a_up
|
||||
! print *,a_lo
|
||||
do tn=N0,Ntime,1
|
||||
! do tn=Ntime,Ntime,1
|
||||
Ntd=tn+1
|
||||
Nt=Ntd-Nd
|
||||
Ntdc=Ntd+Nc
|
||||
indI(4)=Nt
|
||||
indI(5)=Nt+1
|
||||
indI(6)=Nt+2
|
||||
indI(7)=Ntd
|
||||
if (SCIS.gt.0) then
|
||||
if (SCIS.EQ.2) then
|
||||
Nj=max(Nt,0)
|
||||
else
|
||||
Nj=min(max(Nt-5, 0),0)
|
||||
endif
|
||||
endif
|
||||
do ts=3,tn-2
|
||||
!print *,'ts,tn' ,ts,tn,Ntdc
|
||||
CALL COV_INPUT(Big(1:Ntdc,1:Ntdc),tn,ts,R0,R1,R2)!positive wave period
|
||||
indI(2)=ts-2
|
||||
indI(3)=ts-1
|
||||
|
||||
|
||||
CALL RINDD(fxind,Big(1:Ntdc,1:Ntdc),ex(1:Ntdc),
|
||||
& xc,indI,a_lo,a_up)
|
||||
|
||||
ds=dt
|
||||
do icy=1,Nx
|
||||
! ansr(tn,:)=ansr(tn,:)+fxind*CC*ds./(CY1.*CY2)
|
||||
ansr(tn,icy)=ansr(tn,icy)+fxind(icy)*CC*ds/(CY1(icy)*CY2(icy))
|
||||
enddo
|
||||
enddo ! ts
|
||||
print *,'Ready: ',tn,' of ',Ntime
|
||||
|
||||
enddo !tn
|
||||
!print *,'ansr',ansr
|
||||
300 open (unit=11, file='dens.out', STATUS='unknown')
|
||||
!print *, ansr
|
||||
do ts=1,Ntime
|
||||
do ph=1,Nx
|
||||
write(11,*) ansr(ts,ph),hh1(ph),hh2(ph)
|
||||
! write(11,111) ansr(ts,ph)
|
||||
|
||||
enddo
|
||||
enddo
|
||||
!111 FORMAT(2x,F12.8)
|
||||
close(11)
|
||||
900 deallocate(big)
|
||||
deallocate(fxind)
|
||||
deallocate(ansr)
|
||||
deallocate(xc)
|
||||
deallocate(ex)
|
||||
deallocate(R0)
|
||||
deallocate(R1)
|
||||
deallocate(R2)
|
||||
if (allocated(COV) ) then
|
||||
deallocate(COV)
|
||||
endif
|
||||
deallocate(h1)
|
||||
deallocate(h2)
|
||||
deallocate(hh1)
|
||||
deallocate(hh2)
|
||||
deallocate(a_up)
|
||||
deallocate(a_lo)
|
||||
stop
|
||||
!return
|
||||
|
||||
CONTAINS
|
||||
|
||||
|
||||
|
||||
SUBROUTINE INIT_LEVELS
|
||||
& (U,Ntime,N0,NIT,speed,SCIS,seed1,Nx1,Nx2,dT)
|
||||
IMPLICIT NONE
|
||||
integer, intent(out):: Ntime,N0,NIT,speed,Nx1,Nx2,SCIS,seed1
|
||||
double precision ,intent(out) :: U,dT
|
||||
|
||||
|
||||
OPEN(UNIT=14,FILE='reflev.in',STATUS= 'UNKNOWN')
|
||||
READ (14,*) U
|
||||
READ (14,*) Ntime
|
||||
READ (14,*) N0
|
||||
READ (14,*) NIT
|
||||
READ (14,*) speed
|
||||
READ (14,*) SCIS
|
||||
READ (14,*) seed1
|
||||
|
||||
|
||||
READ (14,*) Nx1,Nx2
|
||||
READ (14,*) dT
|
||||
if (Ntime.lt.3) then
|
||||
print *,'The number of wavelength points is too small, stop'
|
||||
stop
|
||||
end if
|
||||
|
||||
CLOSE(UNIT=14)
|
||||
|
||||
RETURN
|
||||
END SUBROUTINE INIT_LEVELS
|
||||
|
||||
C******************************************************
|
||||
SUBROUTINE INIT_AMPLITUDES(h1,Nx1,h2,Nx2)
|
||||
IMPLICIT NONE
|
||||
double precision, dimension(:), intent(out) :: h1,h2
|
||||
integer, intent(in) :: Nx1,Nx2
|
||||
integer :: ix
|
||||
|
||||
|
||||
OPEN(UNIT=4,FILE='h.in',STATUS= 'UNKNOWN')
|
||||
|
||||
C
|
||||
C Reading in amplitudes
|
||||
C
|
||||
do ix=1,Nx1
|
||||
READ (4,*) H1(ix)
|
||||
enddo
|
||||
do ix=1,Nx2
|
||||
READ (4,*) H2(ix)
|
||||
enddo
|
||||
CLOSE(UNIT=4)
|
||||
|
||||
RETURN
|
||||
END SUBROUTINE INIT_AMPLITUDES
|
||||
|
||||
C**************************************************
|
||||
|
||||
C***********************************************************************
|
||||
C***********************************************************************
|
||||
|
||||
SUBROUTINE INIT_COVARIANCES(Ntime,R0,R1,R2)
|
||||
IMPLICIT NONE
|
||||
double precision, dimension(:),intent(out) :: R0,R1,R2
|
||||
integer,intent(in) :: Ntime
|
||||
integer :: i
|
||||
open (unit=1, file='Cd0.in',STATUS='unknown')
|
||||
open (unit=2, file='Cd1.in',STATUS='unknown')
|
||||
open (unit=3, file='Cd2.in',STATUS='unknown')
|
||||
|
||||
do i=1,Ntime
|
||||
read(1,*) R0(i)
|
||||
read(2,*) R1(i)
|
||||
read(3,*) R2(i)
|
||||
enddo
|
||||
close(1)
|
||||
close(2)
|
||||
close(3)
|
||||
|
||||
return
|
||||
END SUBROUTINE INIT_COVARIANCES
|
||||
|
||||
C***********************************************************************
|
||||
C***********************************************************************
|
||||
|
||||
C**********************************************************************
|
||||
|
||||
SUBROUTINE COV_INPUT(BIG,tn,ts, R0,R1,R2)
|
||||
IMPLICIT NONE
|
||||
double precision, dimension(:,:),intent(inout) :: BIG
|
||||
double precision, dimension(:),intent(in) :: R0,R1,R2
|
||||
integer ,intent(in) :: tn,ts
|
||||
integer :: i,j,Ntd1,N !=Ntdc
|
||||
double precision :: tmp
|
||||
! the order of the variables in the covariance matrix
|
||||
! are organized as follows:
|
||||
!
|
||||
! ||X(t2)..X(ts),..X(tn-1)||X'(ts) X'(t1) X'(tn)||Y1 Y2 X(ts) X(t1) X(tn)||
|
||||
! = [Xt Xd Xc]
|
||||
! where
|
||||
!
|
||||
! Xt= time points in the indicator function
|
||||
! Xd= derivatives
|
||||
! Xc=variables to condition on
|
||||
|
||||
! Computations of all covariances follows simple rules: Cov(X(t),X(s))=r(t,s),
|
||||
! then Cov(X'(t),X(s))=dr(t,s)/dt. Now for stationary X(t) we have
|
||||
! a function r(tau) such that Cov(X(t),X(s))=r(s-t) (or r(t-s) will give the same result).
|
||||
!
|
||||
! Consequently Cov(X'(t),X(s)) = -r'(s-t) = -sign(s-t)*r'(|s-t|)
|
||||
! Cov(X'(t),X'(s)) = -r''(s-t) = -r''(|s-t|)
|
||||
! Cov(X''(t),X'(s)) = r'''(s-t) = sign(s-t)*r'''(|s-t|)
|
||||
! Cov(X''(t),X(s)) = r''(s-t) = r''(|s-t|)
|
||||
! Cov(X''(t),X''(s)) = r''''(s-t) = r''''(|s-t|)
|
||||
|
||||
Ntd1=tn+1
|
||||
N=Ntd1+Nc
|
||||
do i=1,tn-2
|
||||
!cov(Xt)
|
||||
do j=i,tn-2
|
||||
BIG(i,j) = R0(j-i+1) ! cov(X(ti+1),X(tj+1))
|
||||
enddo
|
||||
!cov(Xt,Xc)
|
||||
BIG(i ,Ntd1+1) = 0.d0 !cov(X(ti+1),Y1)
|
||||
BIG(i ,Ntd1+2) = 0.d0 !cov(X(ti+1),Y2)
|
||||
BIG(i ,Ntd1+4) = R0(i+1) !cov(X(ti+1),X(t1))
|
||||
BIG(tn-1-i ,Ntd1+5) = R0(i+1) !cov(X(t.. ),X(tn))
|
||||
|
||||
!Cov(Xt,Xd)=cov(X(ti+1),x(tj)
|
||||
BIG(i,Ntd1-1) =-R1(i+1) !cov(X(ti+1),X'(t1))
|
||||
BIG(tn-1-i,Ntd1)= R1(i+1) !cov(X(ti+1),X'(tn))
|
||||
enddo
|
||||
!cov(Xd)
|
||||
BIG(Ntd1 ,Ntd1 ) = -R2(1)
|
||||
BIG(Ntd1-1,Ntd1 ) = -R2(tn) !cov(X'(t1),X'(tn))
|
||||
BIG(Ntd1-1,Ntd1-1) = -R2(1)
|
||||
BIG(Ntd1-2,Ntd1-1) = -R2(ts) !cov(X'(ts),X'(t1))
|
||||
BIG(Ntd1-2,Ntd1-2) = -R2(1)
|
||||
BIG(Ntd1-2,Ntd1 ) = -R2(tn+1-ts) !cov(X'(ts),X'(tn))
|
||||
|
||||
!cov(Xc)
|
||||
BIG(Ntd1+1,Ntd1+1) = 100.d0 ! cov(Y1 Y1)
|
||||
BIG(Ntd1+1,Ntd1+2) = 0.d0 ! cov(Y1 Y2)
|
||||
BIG(Ntd1+1,Ntd1+3) = 0.d0 ! cov(Y1 X(ts))
|
||||
BIG(Ntd1+1,Ntd1+4) = 0.d0 ! cov(Y1 X(t1))
|
||||
BIG(Ntd1+1,Ntd1+5) = 0.d0 ! cov(Y1 X(tn))
|
||||
BIG(Ntd1+2,Ntd1+2) = 100.d0 ! cov(Y2 Y2)
|
||||
BIG(Ntd1+2,Ntd1+3) = 0.d0 ! cov(Y2 X(ts))
|
||||
BIG(Ntd1+2,Ntd1+4) = 0.d0 ! cov(Y2 X(t1))
|
||||
BIG(Ntd1+2,Ntd1+5) = 0.d0 ! cov(Y2 X(tn))
|
||||
|
||||
BIG(Ntd1+3,Ntd1+3) = R0(1) ! cov(X(ts),X (ts)
|
||||
BIG(Ntd1+3,Ntd1+4) = R0(ts) ! cov(X(ts),X (t1))
|
||||
BIG(Ntd1+3,Ntd1+5) = R0(tn+1-ts) ! cov(X(ts),X (tn))
|
||||
BIG(Ntd1+4,Ntd1+4) = R0(1) ! cov(X(t1),X (t1))
|
||||
BIG(Ntd1+4,Ntd1+5) = R0(tn) ! cov(X(t1),X (tn))
|
||||
BIG(Ntd1+5,Ntd1+5) = R0(1) ! cov(X(tn),X (tn))
|
||||
|
||||
|
||||
!cov(Xd,Xc)
|
||||
BIG(Ntd1 ,Ntd1+1) = 0.d0 !cov(X'(tn),Y1)
|
||||
BIG(Ntd1 ,Ntd1+2) = 0.d0 !cov(X'(tn),Y2)
|
||||
BIG(Ntd1-1 ,Ntd1+1) = 0.d0 !cov(X'(t1),Y1)
|
||||
BIG(Ntd1-1 ,Ntd1+2) = 0.d0 !cov(X'(t1),Y2)
|
||||
BIG(Ntd1-2 ,Ntd1+1) = 0.d0 !cov(X'(ts),Y1)
|
||||
BIG(Ntd1-2 ,Ntd1+2) = 0.d0 !cov(X'(ts),Y2)
|
||||
|
||||
BIG(Ntd1 ,Ntd1+4) = R1(tn) !cov(X'(tn),X(t1))
|
||||
BIG(Ntd1 ,Ntd1+5) = 0.d0 !cov(X'(tn),X(tn))
|
||||
BIG(Ntd1-1,Ntd1+4) = 0.d0 !cov(X'(t1),X(t1))
|
||||
BIG(Ntd1-1,Ntd1+5) =-R1(tn) !cov(X'(t1),X(tn))
|
||||
BIG(Ntd1 ,Ntd1+3) = R1(tn+1-ts) !cov(X'(tn),X (ts))
|
||||
BIG(Ntd1-1,Ntd1+3) =-R1(ts) !cov(X'(t1),X (ts))
|
||||
BIG(Ntd1-2,Ntd1+3) = 0.d0 !cov(X'(ts),X (ts)
|
||||
BIG(Ntd1-2,Ntd1+4) = R1(ts) !cov(X'(ts),X (t1))
|
||||
BIG(Ntd1-2,Ntd1+5) = -R1(tn+1-ts) !cov(X'(ts),X (tn))
|
||||
|
||||
|
||||
do i=1,tn-2
|
||||
j=abs(i+1-ts)
|
||||
!cov(Xt,Xc)
|
||||
BIG(i,Ntd1+3) = R0(j+1) !cov(X(ti+1),X(ts))
|
||||
!Cov(Xt,Xd)
|
||||
if ((i+1-ts).lt.0) then
|
||||
BIG(i,Ntd1-2) = R1(j+1)
|
||||
else !cov(X(ti+1),X'(ts))
|
||||
BIG(i,Ntd1-2) = -R1(j+1)
|
||||
endif
|
||||
enddo
|
||||
|
||||
! make lower triangular part equal to upper
|
||||
do j=1,N-1
|
||||
do i=j+1,N
|
||||
tmp =BIG(j,i)
|
||||
|
||||
BIG(i,j)=tmp
|
||||
enddo
|
||||
enddo
|
||||
|
||||
C write (*,10) ((BIG(j,i),i=N+1,N+6),j=N+1,N+6)
|
||||
C 10 format(6F8.4)
|
||||
RETURN
|
||||
END SUBROUTINE COV_INPUT
|
||||
|
||||
SUBROUTINE COV_INPUT2(BIG,pt, R0,R1,R2)
|
||||
IMPLICIT NONE
|
||||
double precision, dimension(:,:), intent(out) :: BIG
|
||||
double precision, dimension(:), intent(in) :: R0,R1,R2
|
||||
integer :: pt,i,j
|
||||
! the order of the variables in the covariance matrix
|
||||
! are organized as follows;
|
||||
! X(t2)...X(tn-1) X'(t1) X'(tn) X(t1) X(tn) = [Xt Xd Xc]
|
||||
!
|
||||
! where Xd is the derivatives
|
||||
!
|
||||
! Xt= time points in the indicator function
|
||||
! Xd= derivatives
|
||||
! Xc=variables to condition on
|
||||
|
||||
!cov(Xc)
|
||||
BIG(pt+2,pt+2) = R0(1)
|
||||
BIG(pt+1,pt+1) = R0(1)
|
||||
BIG(pt+1,pt+2) = R0(pt)
|
||||
!cov(Xd)
|
||||
BIG(pt,pt) = -R2(1)
|
||||
BIG(pt-1,pt-1) = -R2(1)
|
||||
BIG(pt-1,pt) = -R2(pt)
|
||||
!cov(Xd,Xc)
|
||||
BIG(pt,pt+2) = 0.d0
|
||||
BIG(pt,pt+1) = R1(pt)
|
||||
BIG(pt-1,pt+2) = -R1(pt)
|
||||
BIG(pt-1,pt+1) = 0.d0
|
||||
|
||||
if (pt.GT.2) then
|
||||
!cov(Xt)
|
||||
do i=1,pt-2
|
||||
do j=i,pt-2
|
||||
BIG(i,j) = R0(j-i+1)
|
||||
enddo
|
||||
enddo
|
||||
!cov(Xt,Xc)
|
||||
do i=1,pt-2
|
||||
BIG(i,pt+1) = R0(i+1)
|
||||
BIG(pt-1-i,pt+2) = R0(i+1)
|
||||
enddo
|
||||
!Cov(Xt,Xd)=cov(X(ti+1),x(tj))
|
||||
do i=1,pt-2
|
||||
BIG(i,pt-1) = -R1(i+1)
|
||||
BIG(pt-1-i,pt)= R1(i+1)
|
||||
enddo
|
||||
endif
|
||||
|
||||
|
||||
! make lower triangular part equal to upper
|
||||
do j=1,pt+1
|
||||
do i=j+1,pt+2
|
||||
BIG(i,j)=BIG(j,i)
|
||||
enddo
|
||||
enddo
|
||||
C write (*,10) ((BIG(j,i),i=N+1,N+6),j=N+1,N+6)
|
||||
C 10 format(6F8.4)
|
||||
RETURN
|
||||
END SUBROUTINE COV_INPUT2
|
||||
|
||||
|
||||
END PROGRAM sp2tthpdf1
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
@ -1,445 +0,0 @@
|
||||
PROGRAM sp2Acdf
|
||||
C***********************************************************************
|
||||
C This program computes: *
|
||||
C *
|
||||
C density of T_i, for Ac <=h, in a gaussian process i.e. *
|
||||
C *
|
||||
C half wavelength (up-crossing to downcrossing) for crests <h *
|
||||
C or half wavelength (down-crossing to upcrossing) for trough >h *
|
||||
C***********************************************************************
|
||||
use GLOBALDATA, only : Nt,Nj,Nd,Nc,Ntd,Ntdc,NI,Mb,
|
||||
&NIT,Nx,TWOPI,XSPLT,SCIS,NSIMmax,COV
|
||||
use rind
|
||||
IMPLICIT NONE
|
||||
double precision, dimension(:,:),allocatable :: BIG
|
||||
double precision, dimension(:,:),allocatable :: ansr
|
||||
double precision, dimension(: ),allocatable :: ex,CY
|
||||
double precision, dimension(:,:),allocatable :: xc
|
||||
double precision, dimension(: ),allocatable :: fxind,h
|
||||
double precision, dimension(: ),allocatable :: R0,R1,R2,R3,R4
|
||||
double precision ::CC,U,XddInf,XdInf,XtInf
|
||||
double precision, dimension(:,:),allocatable :: a_up,a_lo
|
||||
integer , dimension(: ),allocatable :: seed
|
||||
integer ,dimension(7) :: indI
|
||||
integer :: Nstart,Ntime,tn,ts,speed,ph,def,seed1,seed_size,icy
|
||||
integer ::it1,it2,status
|
||||
double precision :: ds,dT ! lag spacing for covariances
|
||||
! f90 sp2Acdf.f rind51.f
|
||||
|
||||
CALL INIT_LEVELS(U,def,Ntime,Nstart,NIT,speed,Nx,dT)
|
||||
!print *,'U,def,Ntime,Nstart,NIT,speed,SCIS,seed1,Nx,dT'
|
||||
!print *,U,def,Ntime,Nstart,NIT,speed,SCIS,seed1,Nx,dT
|
||||
!XSPLT=1.5d0
|
||||
if (SCIS.GT.0) then
|
||||
allocate(COV(1:Nx))
|
||||
call random_seed(SIZE=seed_size)
|
||||
allocate(seed(seed_size))
|
||||
call random_seed(GET=seed(1:seed_size)) ! get current seed
|
||||
seed(1)=seed1 ! change seed
|
||||
call random_seed(PUT=seed(1:seed_size))
|
||||
deallocate(seed)
|
||||
endif
|
||||
CALL INITDATA(speed)
|
||||
!print *,ntime,speed,u,NIT
|
||||
allocate(R0(1:Ntime+1))
|
||||
allocate(R1(1:Ntime+1))
|
||||
allocate(R2(1:Ntime+1))
|
||||
if (abs(def).GT.1) THEN
|
||||
allocate(R3(1:Ntime+1))
|
||||
allocate(R4(1:Ntime+1))
|
||||
!CALL INIT_AMPLITUDES(h,def,Nx)
|
||||
endif
|
||||
allocate(h(1:Nx))
|
||||
CALL INIT_AMPLITUDES(h,def,Nx)
|
||||
CALL INIT_COVARIANCES(Ntime,def,R0,R1,R2,R3,R4)
|
||||
|
||||
NI=4; Nd=2
|
||||
Nc=3; Mb=2
|
||||
|
||||
Nj=0
|
||||
indI(1)=0
|
||||
C ***** The bound 'infinity' is set to 10*sigma *****
|
||||
XdInf=10.d0*SQRT(-R2(1))
|
||||
XtInf=10.d0*SQRT(R0(1))
|
||||
! normalizing constant
|
||||
CC=TWOPI*SQRT(-R0(1)/R2(1))*exp(u*u/(2.d0*R0(1)) )
|
||||
|
||||
allocate(CY(1:Nx))
|
||||
do icy=1,Nx
|
||||
CY(icy)=exp(-0.5*h(icy)*h(icy)/100)/(10*sqrt(twopi))
|
||||
enddo
|
||||
allocate(BIG(1:Ntime+Nc,1:Ntime+Nc),stat=status)
|
||||
if (status.ne.0) then
|
||||
print *,'can not allocate BIG'
|
||||
end if
|
||||
allocate(ex(1:Ntime+Nc),stat=status)
|
||||
if (status.ne.0) then
|
||||
print *,'can not allocate ex'
|
||||
end if
|
||||
allocate(ansr(1:Ntime,1:Nx))
|
||||
ansr=0.d0
|
||||
allocate(fxind(1:Nx))
|
||||
fxind=0.d0 !this is not needed
|
||||
allocate(xc(1:Nc,1:Nx))
|
||||
|
||||
|
||||
allocate(a_up(Mb,NI-1))
|
||||
allocate(a_lo(Mb,NI-1))
|
||||
a_up=0.d0
|
||||
a_lo=0.d0
|
||||
xc(1,1:Nx)=h(1:Nx)
|
||||
xc(2,1:Nx)=u
|
||||
xc(3,1:Nx)=u
|
||||
|
||||
if (def.GT.0) then
|
||||
a_up(1,1)=0.d0
|
||||
a_lo(1,1)=u
|
||||
a_up(1,2)=XdInf
|
||||
a_lo(1,3)=-XdInf
|
||||
a_up(2,1)=1.d0
|
||||
else
|
||||
a_up(1,1)=u
|
||||
a_lo(1,1)=0.d0
|
||||
a_lo(1,2)=-XdInf
|
||||
a_up(1,3)= XdInf
|
||||
a_lo(2,1)=1.d0
|
||||
endif
|
||||
!print *,'Nstart',Nstart
|
||||
Nstart=MAX(3,Nstart)
|
||||
|
||||
|
||||
if (SCIS.GT.0) then
|
||||
open (unit=11, file='COV.out', STATUS='unknown')
|
||||
write(11,*) 0.d0
|
||||
endif
|
||||
|
||||
!print *,'loop starts'
|
||||
do Ntd=Nstart,Ntime
|
||||
|
||||
Ntdc=Ntd+Nc
|
||||
ex=0.d0
|
||||
BIG=0.d0
|
||||
CALL COV_INPUT(BIG(1:Ntdc,1:Ntdc),Ntd,-1,R0,R1,R2,R3,R4) ! positive wave period
|
||||
C CALL ECHO(BIG(1:2,1:2))
|
||||
Nt=Ntd-Nd;
|
||||
indI(2)=Nt;
|
||||
indI(3)=Nt+1;
|
||||
indI(4)=Ntd;
|
||||
|
||||
CALL RINDD(fxind,Big(1:Ntdc,1:Ntdc),ex,xc,indI,a_lo,a_up)
|
||||
!print *,'test',fxind/CY(1:Nx)
|
||||
|
||||
do icy=1,Nx
|
||||
ansr(Ntd,icy)=fxind(icy)*CC/CY(icy)
|
||||
enddo
|
||||
if (SCIS.GT.0) then
|
||||
write(11,*) COV(1) ! save coefficient of variation
|
||||
endif
|
||||
if((Nx.gt.4).or.NIT.gt.5) print *,'Ready: ',Ntd,' of ',Ntime
|
||||
enddo
|
||||
goto 300
|
||||
300 open (unit=11, file='dens.out', STATUS='unknown')
|
||||
|
||||
!print *, ansr
|
||||
do ts=1,Ntime
|
||||
do ph=1,Nx
|
||||
write(11,*) ansr(ts,ph)
|
||||
enddo
|
||||
enddo
|
||||
!111 FORMAT(2x,F12.8)
|
||||
close(11)
|
||||
900 continue
|
||||
deallocate(BIG)
|
||||
deallocate(ex)
|
||||
deallocate(fxind)
|
||||
deallocate(ansr)
|
||||
deallocate(xc)
|
||||
deallocate(R0)
|
||||
deallocate(R1)
|
||||
deallocate(R2)
|
||||
if (allocated(COV) ) then
|
||||
deallocate(COV)
|
||||
endif
|
||||
|
||||
if (allocated(R3)) then
|
||||
deallocate(R3)
|
||||
deallocate(R4)
|
||||
deallocate(h)
|
||||
ENDIF
|
||||
stop
|
||||
!return
|
||||
|
||||
CONTAINS
|
||||
|
||||
|
||||
|
||||
SUBROUTINE INIT_LEVELS
|
||||
& (U,def,Ntime,Nstart,NIT,speed,Nx,dT)
|
||||
IMPLICIT NONE
|
||||
integer, intent(out):: def,Ntime,Nstart,NIT,speed,Nx
|
||||
double precision ,intent(out) :: U,dT
|
||||
|
||||
|
||||
OPEN(UNIT=14,FILE='reflev.in',STATUS= 'UNKNOWN')
|
||||
READ (14,*) U
|
||||
READ (14,*) def
|
||||
READ (14,*) Ntime
|
||||
READ (14,*) Nstart
|
||||
READ (14,*) NIT
|
||||
READ (14,*) speed
|
||||
READ (14,*) SCIS
|
||||
READ (14,*) seed1
|
||||
READ (14,*) Nx
|
||||
|
||||
if (abs(def).GT.1) then
|
||||
READ (14,*) dT
|
||||
if (Ntime.lt.3) then
|
||||
print *,'The number of wavelength points is too small, stop'
|
||||
stop
|
||||
end if
|
||||
else
|
||||
if (Ntime.lt.2) then
|
||||
print *,'The number of wavelength points is too small, stop'
|
||||
stop
|
||||
end if
|
||||
endif
|
||||
CLOSE(UNIT=14)
|
||||
|
||||
RETURN
|
||||
END SUBROUTINE INIT_LEVELS
|
||||
|
||||
C******************************************************
|
||||
SUBROUTINE INIT_AMPLITUDES(h,def,Nx)
|
||||
IMPLICIT NONE
|
||||
double precision, dimension(:), intent(out) :: h
|
||||
integer, intent(in) :: def
|
||||
integer, intent(in) :: Nx
|
||||
integer :: ix
|
||||
|
||||
|
||||
OPEN(UNIT=4,FILE='h.in',STATUS= 'UNKNOWN')
|
||||
|
||||
C
|
||||
C Reading in amplitudes
|
||||
C
|
||||
do ix=1,Nx
|
||||
READ (4,*) H(ix)
|
||||
enddo
|
||||
CLOSE(UNIT=4)
|
||||
!if (def.LT.0) THEN
|
||||
! H=-H
|
||||
!endif
|
||||
|
||||
RETURN
|
||||
END SUBROUTINE INIT_AMPLITUDES
|
||||
|
||||
C**************************************************
|
||||
|
||||
C***********************************************************************
|
||||
C***********************************************************************
|
||||
|
||||
SUBROUTINE INIT_COVARIANCES(Ntime,def,R0,R1,R2,R3,R4)
|
||||
IMPLICIT NONE
|
||||
double precision, dimension(:),intent(out) :: R0,R1,R2
|
||||
double precision, dimension(:),intent(out) :: R3,R4
|
||||
integer,intent(in) :: Ntime,def
|
||||
integer :: i
|
||||
open (unit=1, file='Cd0.in',STATUS='unknown')
|
||||
open (unit=2, file='Cd1.in',STATUS='unknown')
|
||||
open (unit=3, file='Cd2.in',STATUS='unknown')
|
||||
|
||||
do i=1,Ntime
|
||||
read(1,*) R0(i)
|
||||
read(2,*) R1(i)
|
||||
read(3,*) R2(i)
|
||||
enddo
|
||||
close(1)
|
||||
close(2)
|
||||
close(3)
|
||||
|
||||
if (abs(def).GT.1) then
|
||||
open (unit=4, file='Cd3.in',STATUS='unknown')
|
||||
open (unit=5, file='Cd4.in',STATUS='unknown')
|
||||
|
||||
do i=1,Ntime
|
||||
read(4,*) R3(i)
|
||||
read(5,*) R4(i)
|
||||
enddo
|
||||
|
||||
close(4)
|
||||
close(5)
|
||||
endif
|
||||
return
|
||||
END SUBROUTINE INIT_COVARIANCES
|
||||
|
||||
C***********************************************************************
|
||||
C***********************************************************************
|
||||
|
||||
C**********************************************************************
|
||||
|
||||
SUBROUTINE COV_INPUT(BIG,tn,ts, R0,R1,R2,R3,R4)
|
||||
IMPLICIT NONE
|
||||
double precision, dimension(:,:),intent(inout) :: BIG
|
||||
double precision, dimension(:),intent(in) :: R0,R1,R2
|
||||
double precision, dimension(:),intent(in) :: R3,R4
|
||||
integer ,intent(in) :: tn,ts
|
||||
integer :: i,j,shft,Ntd1,N !=Ntdc
|
||||
double precision :: tmp
|
||||
! the order of the variables in the covariance matrix
|
||||
! are organized as follows:
|
||||
! For ts>1:
|
||||
! X(t2)..X(ts),..X(tn-1) X''(ts) X'(t1) X'(tn) X(ts) X(t1) X(tn) X'(ts)
|
||||
! = [Xt Xd Xc]
|
||||
!
|
||||
! For ts<=1:
|
||||
! X(t2)..,..X(tn-1) X'(t1) X'(tn) Y X(t1) X(tn)
|
||||
! = [Xt Xd Xc]
|
||||
!Add Y Condition : Y=h
|
||||
|
||||
! where
|
||||
!
|
||||
! Xt= time points in the indicator function
|
||||
! Xd= derivatives
|
||||
! Xc=variables to condition on
|
||||
|
||||
if (ts.LE.1) THEN
|
||||
Ntd1=tn
|
||||
N=Ntd1+Nc;
|
||||
shft=0 ! def=1 want only crest period Tc
|
||||
else
|
||||
Ntd1=tn+1
|
||||
N=Ntd1+4
|
||||
shft=1 ! def=2 or 3 want Tc Ac or Tcf, Ac
|
||||
endif
|
||||
|
||||
do i=1,tn-2
|
||||
!cov(Xt)
|
||||
do j=i,tn-2
|
||||
BIG(i,j) = R0(j-i+1) ! cov(X(ti+1),X(tj+1))
|
||||
enddo
|
||||
!cov(Xt,Xc)
|
||||
BIG(i ,Ntd1+1+shft) = 0.d0 !cov(X(ti+1),Y)
|
||||
BIG(i ,Ntd1+2+shft) = R0(i+1) !cov(X(ti+1),X(t1))
|
||||
BIG(tn-1-i ,Ntd1+3+shft) = R0(i+1) !cov(X(t.. ),X(tn))
|
||||
!Cov(Xt,Xd)=cov(X(ti+1),x(tj)
|
||||
BIG(i,Ntd1-1) =-R1(i+1) !cov(X(ti+1),X' (t1))
|
||||
BIG(tn-1-i,Ntd1)= R1(i+1) !cov(X(ti+1),X' (tn))
|
||||
enddo
|
||||
!call echo(big(1:tn,1:tn),tn)
|
||||
!cov(Xd)
|
||||
BIG(Ntd1 ,Ntd1 ) = -R2(1)
|
||||
BIG(Ntd1-1,Ntd1 ) = -R2(tn) !cov(X'(t1),X'(tn))
|
||||
BIG(Ntd1-1,Ntd1-1) = -R2(1)
|
||||
|
||||
!cov(Xc)
|
||||
!print *,'t'
|
||||
BIG(Ntd1+1+shft,Ntd1+1+shft) = 100.d0!100.d0 ! cov(Y,Y)
|
||||
BIG(Ntd1+1+shft,Ntd1+2+shft) = 0.d0
|
||||
BIG(Ntd1+1+shft,Ntd1+3+shft) = 0.d0
|
||||
BIG(Ntd1+2+shft,Ntd1+2+shft) = R0(1) ! cov(X(t1),X (t1))
|
||||
BIG(Ntd1+2+shft,Ntd1+3+shft) = R0(tn) ! cov(X(t1),X (tn))
|
||||
BIG(Ntd1+3+shft,Ntd1+3+shft) = R0(1) ! cov(X(tn),X (tn))
|
||||
!cov(Xd,Xc)
|
||||
BIG(Ntd1 ,Ntd1+1+shft) = 0.d0 !cov(X'(tn),Y)
|
||||
BIG(Ntd1 ,Ntd1+2+shft) = R1(tn) !cov(X'(tn),X(t1))
|
||||
BIG(Ntd1 ,Ntd1+3+shft) = 0.d0 !cov(X'(tn),X(tn))
|
||||
BIG(Ntd1-1,Ntd1+1+shft) = 0.d0 !cov(X'(t1),Y)
|
||||
BIG(Ntd1-1,Ntd1+2+shft) = 0.d0 !cov(X'(t1),X(t1))
|
||||
BIG(Ntd1-1,Ntd1+3+shft) =-R1(tn) !cov(X'(t1),X(tn))
|
||||
|
||||
|
||||
!call echo(big(1:N,1:N),N)
|
||||
! make lower triangular part equal to upper
|
||||
do j=1,N-1
|
||||
do i=j+1,N
|
||||
tmp =BIG(j,i)
|
||||
BIG(i,j)=tmp
|
||||
enddo
|
||||
!call echo(big(1:N,1:N),N)
|
||||
enddo
|
||||
!if (tn.eq.3) then
|
||||
!do j=1,N
|
||||
! do i=j,N
|
||||
! print *,'test',j,i,BIG(j,i)
|
||||
! enddo
|
||||
!call echo(big(1:N,1:N),N)
|
||||
!enddo
|
||||
!endif
|
||||
!call echo(big(1:N,1:N),N)
|
||||
|
||||
C write (*,10) ((BIG(j,i),i=N+1,N+6),j=N+1,N+6)
|
||||
C 10 format(6F8.4)
|
||||
RETURN
|
||||
END SUBROUTINE COV_INPUT
|
||||
|
||||
SUBROUTINE COV_INPUT2(BIG,pt, R0,R1,R2)
|
||||
IMPLICIT NONE
|
||||
double precision, dimension(:,:), intent(out) :: BIG
|
||||
double precision, dimension(:), intent(in) :: R0,R1,R2
|
||||
integer :: pt,i,j
|
||||
! the order of the variables in the covariance matrix
|
||||
! are organized as follows;
|
||||
! X(t2)...X(tn-1) X'(t1) X'(tn) X(t1) X(tn) = [Xt Xd Xc]
|
||||
!
|
||||
! where Xd is the derivatives
|
||||
!
|
||||
! Xt= time points in the indicator function
|
||||
! Xd= derivatives
|
||||
! Xc=variables to condition on
|
||||
|
||||
!cov(Xc)
|
||||
BIG(pt+2,pt+2) = R0(1)
|
||||
BIG(pt+1,pt+1) = R0(1)
|
||||
BIG(pt+1,pt+2) = R0(pt)
|
||||
!cov(Xd)
|
||||
BIG(pt,pt) = -R2(1)
|
||||
BIG(pt-1,pt-1) = -R2(1)
|
||||
BIG(pt-1,pt) = -R2(pt)
|
||||
!cov(Xd,Xc)
|
||||
BIG(pt,pt+2) = 0.d0
|
||||
BIG(pt,pt+1) = R1(pt)
|
||||
BIG(pt-1,pt+2) = -R1(pt)
|
||||
BIG(pt-1,pt+1) = 0.d0
|
||||
|
||||
if (pt.GT.2) then
|
||||
!cov(Xt)
|
||||
do i=1,pt-2
|
||||
do j=i,pt-2
|
||||
BIG(i,j) = R0(j-i+1)
|
||||
enddo
|
||||
enddo
|
||||
!cov(Xt,Xc)
|
||||
do i=1,pt-2
|
||||
BIG(i,pt+1) = R0(i+1)
|
||||
BIG(pt-1-i,pt+2) = R0(i+1)
|
||||
enddo
|
||||
!Cov(Xt,Xd)=cov(X(ti+1),x(tj))
|
||||
do i=1,pt-2
|
||||
BIG(i,pt-1) = -R1(i+1)
|
||||
BIG(pt-1-i,pt)= R1(i+1)
|
||||
enddo
|
||||
endif
|
||||
|
||||
|
||||
! make lower triangular part equal to upper
|
||||
do j=1,pt+1
|
||||
do i=j+1,pt+2
|
||||
BIG(i,j)=BIG(j,i)
|
||||
enddo
|
||||
enddo
|
||||
C write (*,10) ((BIG(j,i),i=N+1,N+6),j=N+1,N+6)
|
||||
C 10 format(6F8.4)
|
||||
RETURN
|
||||
END SUBROUTINE COV_INPUT2
|
||||
|
||||
|
||||
END PROGRAM sp2Acdf
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
@ -1,357 +0,0 @@
|
||||
PROGRAM cov2mmpdf
|
||||
C*******************************************************************************
|
||||
C This program computes joint density of maximum and the following minimum *
|
||||
C*******************************************************************************
|
||||
use GLOBALDATA, only : Nt,Nj,Nd,Nc,Ntd,Ntdc,NI,Mb,
|
||||
&NIT,Nx,TWOPI,XSPLT,SCIS,NSIMmax,COV
|
||||
use rind
|
||||
IMPLICIT NONE
|
||||
double precision, dimension(:,:),allocatable :: BIG
|
||||
double precision, dimension(:,:),allocatable :: ansr
|
||||
double precision, dimension(: ),allocatable :: ex
|
||||
double precision, dimension(:,:),allocatable :: xc
|
||||
double precision, dimension(: ),allocatable :: fxind,h
|
||||
double precision, dimension(: ),allocatable :: R0,R1,R2,R3,R4
|
||||
double precision ::CC,U,XddInf,XdInf,XtInf
|
||||
double precision, dimension(:,:),allocatable :: a_up,a_lo
|
||||
integer , dimension(: ),allocatable :: seed
|
||||
integer ,dimension(7) :: indI
|
||||
integer :: Nstart,Ntime,tn,ts,speed,seed1,seed_size
|
||||
integer :: status,i,j,ij,Nx1
|
||||
double precision :: ds,dT ! lag spacing for covariances
|
||||
|
||||
! f90 cov2mmpdf.f rind51.f
|
||||
|
||||
CALL INIT_LEVELS(Ntime,Nstart,NIT,speed,Nx1,dT)
|
||||
Nx=Nx1*(Nx1-1)/2
|
||||
if (SCIS.GT.0) then
|
||||
allocate(COV(1:Nx))
|
||||
call random_seed(SIZE=seed_size)
|
||||
allocate(seed(seed_size))
|
||||
call random_seed(GET=seed(1:seed_size)) ! get current seed
|
||||
seed(1)=seed1 ! change seed
|
||||
call random_seed(PUT=seed(1:seed_size))
|
||||
deallocate(seed)
|
||||
endif
|
||||
|
||||
CALL INITDATA(speed)
|
||||
|
||||
allocate(R0(1:Ntime+1))
|
||||
allocate(R1(1:Ntime+1))
|
||||
allocate(R2(1:Ntime+1))
|
||||
allocate(R3(1:Ntime+1))
|
||||
allocate(R4(1:Ntime+1))
|
||||
allocate(h(1:Nx1))
|
||||
|
||||
CALL INIT_AMPLITUDES(h,Nx1)
|
||||
CALL INIT_COVARIANCES(Ntime,R0,R1,R2,R3,R4)
|
||||
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
|
||||
! Y= X'(t2)...X'(tn-1)||X''(t1) X''(tn)|| X'(t1) X'(tn) X(t1) X(tn) !!
|
||||
! = [ Xt Xd Xc ] !!
|
||||
! !!
|
||||
! Nt=tn-2, Nd=2, Nc=4 !!
|
||||
! !!
|
||||
! Xt= contains Nt time points in the indicator function !!
|
||||
! Xd= " Nd derivatives !!
|
||||
! Xc= " Nc variables to condition on !!
|
||||
! !!
|
||||
! There are 3 ( NI=4) regions with constant bariers: !!
|
||||
! (indI(1)=0); for i\in (indI(1),indI(2)] Y(i)<0. !!
|
||||
! (indI(2)=Nt) ; for i\in (indI(2)+1,indI(3)], Y(i)<0 (deriv. X''(t1)) !!
|
||||
! (indI(3)=Nt+1); for i\in (indI(3)+1,indI(4)], Y(i)>0 (deriv. X''(tn)) !!
|
||||
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
|
||||
|
||||
|
||||
NI=4; Nd=2
|
||||
Nc=4; Mb=1
|
||||
|
||||
Nj=0
|
||||
indI(1)=0
|
||||
C ***** The bound 'infinity' is set to 10*sigma *****
|
||||
XdInf=10.d0*SQRT(R4(1))
|
||||
XtInf=10.d0*SQRT(-R2(1))
|
||||
! normalizing constant
|
||||
CC=TWOPI*SQRT(-R2(1)/R4(1))
|
||||
|
||||
allocate(BIG(1:Ntime+Nc,1:Ntime+Nc),stat=status)
|
||||
if (status.ne.0) then
|
||||
print *,'can not allocate BIG'
|
||||
end if
|
||||
allocate(ex(1:Ntime+Nc),stat=status)
|
||||
if (status.ne.0) then
|
||||
print *,'can not allocate ex'
|
||||
end if
|
||||
if (Nx.gt.1) then
|
||||
allocate(ansr(1:Nx1,1:Nx1))
|
||||
else
|
||||
allocate(ansr(1,1:Ntime))
|
||||
end if
|
||||
ansr=0.d0
|
||||
allocate(fxind(1:Nx))
|
||||
fxind=0.d0 !this is not needed
|
||||
allocate(xc(1:Nc,1:Nx))
|
||||
|
||||
|
||||
allocate(a_up(Mb,NI-1))
|
||||
allocate(a_lo(Mb,NI-1))
|
||||
|
||||
a_up=0.d0
|
||||
a_lo=0.d0
|
||||
|
||||
ij=0
|
||||
do i=2,Nx1
|
||||
do j=1,i-1
|
||||
ij=ij+1
|
||||
xc(3,ij)=h(i)
|
||||
xc(4,ij)=h(j)
|
||||
enddo
|
||||
enddo
|
||||
xc(1,1:Nx)=0.d0
|
||||
xc(2,1:Nx)=0.d0
|
||||
|
||||
a_lo(1,1)=-Xtinf
|
||||
a_lo(1,2)=-XdInf
|
||||
a_up(1,3)=+XdInf
|
||||
|
||||
|
||||
Nstart=MAX(2,Nstart)
|
||||
|
||||
|
||||
if (SCIS.GT.0) then
|
||||
open (unit=11, file='COV.out', STATUS='unknown')
|
||||
write(11,*) 0.d0
|
||||
endif
|
||||
|
||||
do Ntd=Nstart,Ntime
|
||||
|
||||
Ntdc=Ntd+Nc
|
||||
ex=0.d0
|
||||
BIG=0.d0
|
||||
CALL COV_INPUT(BIG(1:Ntdc,1:Ntdc),Ntd,R0,R1,R2,R3,R4) ! positive wave period
|
||||
|
||||
Nt=Ntd-Nd;
|
||||
indI(2)=Nt;
|
||||
indI(3)=Nt+1;
|
||||
indI(4)=Ntd;
|
||||
|
||||
CALL RINDD(fxind,Big(1:Ntdc,1:Ntdc),ex,xc,indI,a_lo,a_up)
|
||||
ij=0
|
||||
if (Nx .gt. 1) then
|
||||
do i=2,Nx1
|
||||
do j=1,i-1
|
||||
ij=ij+1
|
||||
ansr(i,j)=ansr(i,j)+fxind(ij)*CC*dt
|
||||
enddo
|
||||
enddo
|
||||
else
|
||||
ansr(1,Ntd)=fxind(1)*CC
|
||||
end if
|
||||
|
||||
if (SCIS.GT.0) then
|
||||
write(11,*) COV(1) ! save coefficient of variation
|
||||
endif
|
||||
print *,'Ready: ',Ntd,' of ',Ntime
|
||||
enddo
|
||||
goto 300
|
||||
300 open (unit=11, file='dens.out', STATUS='unknown')
|
||||
if (Nx.gt.1) then
|
||||
do i=1,Nx1
|
||||
do j=1,Nx1
|
||||
write(11,*) ansr(i,j)
|
||||
enddo
|
||||
enddo
|
||||
else
|
||||
do j=1,Ntime
|
||||
write(11,*) ansr(1,j)
|
||||
enddo
|
||||
end if
|
||||
close(11)
|
||||
900 continue
|
||||
deallocate(BIG)
|
||||
deallocate(ex)
|
||||
deallocate(fxind)
|
||||
deallocate(ansr)
|
||||
deallocate(xc)
|
||||
deallocate(R0)
|
||||
deallocate(R1)
|
||||
deallocate(R2)
|
||||
deallocate(R3)
|
||||
deallocate(R4)
|
||||
deallocate(h)
|
||||
|
||||
if (allocated(COV) ) then
|
||||
deallocate(COV)
|
||||
endif
|
||||
stop
|
||||
!return
|
||||
|
||||
CONTAINS
|
||||
|
||||
|
||||
|
||||
SUBROUTINE INIT_LEVELS
|
||||
& (Ntime,Nstart,NIT,speed,Nx,dT)
|
||||
IMPLICIT NONE
|
||||
integer, intent(out):: Ntime,Nstart,NIT,speed,Nx
|
||||
double precision ,intent(out) :: dT
|
||||
|
||||
|
||||
OPEN(UNIT=14,FILE='reflev.in',STATUS= 'UNKNOWN')
|
||||
READ (14,*) Ntime
|
||||
READ (14,*) Nstart
|
||||
READ (14,*) NIT
|
||||
READ (14,*) speed
|
||||
READ (14,*) SCIS
|
||||
READ (14,*) seed1
|
||||
READ (14,*) Nx
|
||||
READ (14,*) dT
|
||||
|
||||
if (Ntime.lt.2) then
|
||||
print *,'The number of wavelength points is too small, stop'
|
||||
stop
|
||||
end if
|
||||
CLOSE(UNIT=14)
|
||||
|
||||
RETURN
|
||||
END SUBROUTINE INIT_LEVELS
|
||||
|
||||
C******************************************************
|
||||
SUBROUTINE INIT_AMPLITUDES(h,Nx)
|
||||
IMPLICIT NONE
|
||||
double precision, dimension(:), intent(out) :: h
|
||||
integer, intent(in) :: Nx
|
||||
integer :: ix
|
||||
|
||||
|
||||
OPEN(UNIT=4,FILE='h.in',STATUS= 'UNKNOWN')
|
||||
|
||||
C
|
||||
C Reading in amplitudes
|
||||
C
|
||||
do ix=1,Nx
|
||||
READ (4,*) H(ix)
|
||||
enddo
|
||||
CLOSE(UNIT=4)
|
||||
|
||||
RETURN
|
||||
END SUBROUTINE INIT_AMPLITUDES
|
||||
|
||||
C**************************************************
|
||||
|
||||
C***********************************************************************
|
||||
C***********************************************************************
|
||||
|
||||
SUBROUTINE INIT_COVARIANCES(Ntime,R0,R1,R2,R3,R4)
|
||||
IMPLICIT NONE
|
||||
double precision, dimension(:),intent(out) :: R0,R1,R2
|
||||
double precision, dimension(:),intent(out) :: R3,R4
|
||||
integer,intent(in) :: Ntime
|
||||
integer :: i
|
||||
open (unit=1, file='Cd0.in',STATUS='unknown')
|
||||
open (unit=2, file='Cd1.in',STATUS='unknown')
|
||||
open (unit=3, file='Cd2.in',STATUS='unknown')
|
||||
open (unit=4, file='Cd3.in',STATUS='unknown')
|
||||
open (unit=5, file='Cd4.in',STATUS='unknown')
|
||||
|
||||
do i=1,Ntime
|
||||
read(1,*) R0(i)
|
||||
read(2,*) R1(i)
|
||||
read(3,*) R2(i)
|
||||
read(4,*) R3(i)
|
||||
read(5,*) R4(i)
|
||||
enddo
|
||||
close(1)
|
||||
close(2)
|
||||
close(3)
|
||||
close(3)
|
||||
close(5)
|
||||
return
|
||||
END SUBROUTINE INIT_COVARIANCES
|
||||
|
||||
C**********************************************************************
|
||||
|
||||
SUBROUTINE COV_INPUT(BIG,tn,R0,R1,R2,R3,R4)
|
||||
IMPLICIT NONE
|
||||
double precision, dimension(:,:),intent(inout) :: BIG
|
||||
double precision, dimension(:),intent(in) :: R0,R1,R2
|
||||
double precision, dimension(:),intent(in) :: R3,R4
|
||||
integer ,intent(in) :: tn
|
||||
integer :: i,j,N
|
||||
double precision :: tmp
|
||||
! the order of the variables in the covariance matrix
|
||||
! are organized as follows:
|
||||
! X'(t2)..X'(ts),...,X'(tn-1) X''(t1),X''(tn) X'(t1),X'(tn),X(t1),X(tn)
|
||||
! = [ Xt | Xd | Xc ]
|
||||
!
|
||||
! where
|
||||
!
|
||||
! Xt= time points in the indicator function
|
||||
! Xd= derivatives
|
||||
! Xc=variables to condition on
|
||||
|
||||
! Computations of all covariances follows simple rules: Cov(X(t),X(s))=r(t,s),
|
||||
! then Cov(X'(t),X(s))=dr(t,s)/dt. Now for stationary X(t) we have
|
||||
! a function r(tau) such that Cov(X(t),X(s))=r(s-t) (or r(t-s) will give the same result).
|
||||
!
|
||||
! Consequently Cov(X'(t),X(s)) = -r'(s-t) = -sign(s-t)*r'(|s-t|)
|
||||
! Cov(X'(t),X'(s)) = -r''(s-t) = -r''(|s-t|)
|
||||
! Cov(X''(t),X'(s)) = r'''(s-t) = sign(s-t)*r'''(|s-t|)
|
||||
! Cov(X''(t),X(s)) = r''(s-t) = r''(|s-t|)
|
||||
! Cov(X''(t),X''(s)) = r''''(s-t) = r''''(|s-t|)
|
||||
|
||||
N=tn+4
|
||||
do i=1,tn-2
|
||||
!cov(Xt)
|
||||
do j=i,tn-2
|
||||
BIG(i,j) = -R2(j-i+1) ! cov(X'(ti+1),X'(tj+1))
|
||||
enddo
|
||||
!cov(Xt,Xc)
|
||||
BIG(i ,tn+3) = R1(i+1) !cov(X'(ti+1),X(t1))
|
||||
BIG(tn-1-i ,tn+4) = -R1(i+1) !cov(X'(ti+1),X(tn))
|
||||
BIG(i ,tn+1) = -R2(i+1) !cov(X'(ti+1),X'(t1))
|
||||
BIG(tn-1-i ,tn+2) = -R2(i+1) !cov(X'(ti+1),X'(tn))
|
||||
!Cov(Xt,Xd)
|
||||
BIG(i,tn-1) = R3(i+1) !cov(X'(ti+1),X''(t1))
|
||||
BIG(tn-1-i,tn) =-R3(i+1) !cov(X'(ti+1),X''(tn))
|
||||
enddo
|
||||
|
||||
!cov(Xd)
|
||||
BIG(tn-1 ,tn-1 ) = R4(1)
|
||||
BIG(tn-1,tn ) = R4(tn) !cov(X''(t1),X''(tn))
|
||||
BIG(tn ,tn ) = R4(1)
|
||||
|
||||
!cov(Xc)
|
||||
BIG(tn+3,tn+3) = R0(1) ! cov(X(t1),X(t1))
|
||||
BIG(tn+3,tn+4) = R0(tn) ! cov(X(t1),X(tn))
|
||||
BIG(tn+1,tn+3) = 0.d0 ! cov(X(t1),X'(t1))
|
||||
BIG(tn+2,tn+3) = R1(tn) ! cov(X(t1),X'(tn))
|
||||
BIG(tn+4,tn+4) = R0(1) ! cov(X(tn),X(tn))
|
||||
BIG(tn+1,tn+4) =-R1(tn) ! cov(X(tn),X'(t1))
|
||||
BIG(tn+2,tn+4) = 0.d0 ! cov(X(tn),X'(tn))
|
||||
BIG(tn+1,tn+1) =-R2(1) ! cov(X'(t1),X'(t1))
|
||||
BIG(tn+1,tn+2) =-R2(tn) ! cov(X'(t1),X'(tn))
|
||||
BIG(tn+2,tn+2) =-R2(1) ! cov(X'(tn),X'(tn))
|
||||
!Xc=X(t1),X(tn),X'(t1),X'(tn)
|
||||
!Xd=X''(t1),X''(tn)
|
||||
!cov(Xd,Xc)
|
||||
BIG(tn-1 ,tn+3) = R2(1) !cov(X''(t1),X(t1))
|
||||
BIG(tn-1 ,tn+4) = R2(tn) !cov(X''(t1),X(tn))
|
||||
BIG(tn-1 ,tn+1) = 0.d0 !cov(X''(t1),X'(t1))
|
||||
BIG(tn-1 ,tn+2) = R3(tn) !cov(X''(t1),X'(tn))
|
||||
BIG(tn ,tn+3) = R2(tn) !cov(X''(tn),X(t1))
|
||||
BIG(tn ,tn+4) = R2(1) !cov(X''(tn),X(tn))
|
||||
BIG(tn ,tn+1) =-R3(tn) !cov(X''(tn),X'(t1))
|
||||
BIG(tn ,tn+2) = 0.d0 !cov(X''(tn),X'(tn))
|
||||
! make lower triangular part equal to upper
|
||||
do j=1,N-1
|
||||
do i=j+1,N
|
||||
tmp =BIG(j,i)
|
||||
BIG(i,j)=tmp
|
||||
enddo
|
||||
enddo
|
||||
RETURN
|
||||
END SUBROUTINE COV_INPUT
|
||||
|
||||
|
||||
END PROGRAM cov2mmpdf
|
||||
|
@ -1,769 +0,0 @@
|
||||
PROGRAM sp2mmt
|
||||
C*******************************************************************************
|
||||
C This program computes joint density of the maximum and the following *
|
||||
C minimum or level u separated maxima and minima + period/wavelength *
|
||||
C*******************************************************************************
|
||||
use GLOBALDATA, only : Nt,Nj,Nd,Nc,Ntd,Ntdc,NI,Mb,
|
||||
&NIT,Nx,TWOPI,XSPLT,SCIS,NSIMmax,COV
|
||||
use rind
|
||||
IMPLICIT NONE
|
||||
double precision, dimension(:,:), allocatable :: BIG
|
||||
double precision, dimension(:,:,:),allocatable :: ansr
|
||||
double precision, dimension(: ), allocatable :: ex
|
||||
double precision, dimension(:,:), allocatable :: xc
|
||||
double precision, dimension(: ), allocatable :: fxind,h
|
||||
double precision, dimension(: ), allocatable :: R0,R1,R2,R3,R4
|
||||
double precision :: CC,U,XdInf,XtInf
|
||||
double precision, dimension(1,4) :: a_up,a_lo ! size Mb X NI-1
|
||||
integer , dimension(: ), allocatable :: seed
|
||||
integer ,dimension(5) :: indI = 0 ! length NI
|
||||
integer :: Nstart,Ntime,ts,tn,speed,seed1,seed_size
|
||||
integer :: status,i,j,ij,Nx0,Nx1,DEF,isOdd !,TMP
|
||||
LOGICAL :: SYMMETRY=.FALSE.
|
||||
double precision :: dT ! lag spacing for covariances
|
||||
|
||||
! f90 -gline -fieee -Nl126 -C -o intmodule.f rind60.f sp2mmt.f
|
||||
|
||||
CALL INIT_LEVELS(Ntime,Nstart,NIT,speed,SCIS,SEED1,Nx1,dT,u,def)
|
||||
CALL INITDATA(speed)
|
||||
|
||||
if (SCIS.GT.0) then
|
||||
!allocate(COV(1:Nx))
|
||||
call random_seed(SIZE=seed_size)
|
||||
allocate(seed(seed_size))
|
||||
call random_seed(GET=seed(1:seed_size)) ! get current seed
|
||||
seed(1)=seed1 ! change seed
|
||||
call random_seed(PUT=seed(1:seed_size))
|
||||
deallocate(seed)
|
||||
if (ALLOCATED(COV)) then
|
||||
open (unit=11, file='COV.out', STATUS='unknown')
|
||||
write(11,*) 0.d0
|
||||
endif
|
||||
endif
|
||||
|
||||
allocate(R0(1:Ntime+1))
|
||||
allocate(R1(1:Ntime+1))
|
||||
allocate(R2(1:Ntime+1))
|
||||
allocate(R3(1:Ntime+1))
|
||||
allocate(R4(1:Ntime+1))
|
||||
|
||||
Nx0 = Nx1 ! just plain Mm
|
||||
IF (def.GT.1) Nx0=2*Nx1 ! level v separated max2min densities wanted
|
||||
|
||||
|
||||
allocate(h(1:Nx0))
|
||||
|
||||
CALL INIT_AMPLITUDES(h,Nx0)
|
||||
CALL INIT_COVARIANCES(Ntime,R0,R1,R2,R3,R4)
|
||||
! For DEF = 0,1 : (Maxima, Minima and period/wavelength)
|
||||
! = 2,3 : (Level v separated Maxima and Minima and period/wavelength between them)
|
||||
! If Nx==1 then the conditional density for period/wavelength between Maxima and Minima
|
||||
! given the Max and Min is returned
|
||||
!~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
|
||||
! Y= X'(t2)..X'(ts)..X'(tn-1)||X''(t1) X''(tn)|| X'(t1) X'(tn) X(t1) X(tn)
|
||||
! = [ Xt Xd Xc ]
|
||||
!
|
||||
! Nt = tn-2, Nd = 2, Nc = 4
|
||||
!
|
||||
! Xt= contains Nt time points in the indicator function
|
||||
! Xd= " Nd derivatives in Jacobian
|
||||
! Xc= " Nc variables to condition on
|
||||
!
|
||||
! There are 3 (NI=4) regions with constant barriers:
|
||||
! (indI(1)=0); for i\in (indI(1),indI(2)] Y(i)<0.
|
||||
! (indI(2)=Nt) ; for i\in (indI(2)+1,indI(3)], Y(i)<0 (deriv. X''(t1))
|
||||
! (indI(3)=Nt+1); for i\in (indI(3)+1,indI(4)], Y(i)>0 (deriv. X''(tn))
|
||||
!
|
||||
!
|
||||
! For DEF = 4,5 (Level v separated Maxima and Minima and period/wavelength from Max to crossing)
|
||||
! If Nx==1 then the conditional joint density for period/wavelength between Maxima, Minima and Max to
|
||||
! level v crossing given the Max and the min is returned
|
||||
!~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
|
||||
! Y= X'(t2)..X'(ts)..X'(tn-1)||X''(t1) X''(tn) X'(ts)|| X'(t1) X'(tn) X(t1) X(tn) X(ts)
|
||||
! = [ Xt Xd Xc ]
|
||||
!
|
||||
! Nt = tn-2, Nd = 3, Nc = 5
|
||||
!
|
||||
! Xt= contains Nt time points in the indicator function
|
||||
! Xd= " Nd derivatives
|
||||
! Xc= " Nc variables to condition on
|
||||
!
|
||||
! There are 4 (NI=5) regions with constant barriers:
|
||||
! (indI(1)=0); for i\in (indI(1),indI(2)] Y(i)<0.
|
||||
! (indI(2)=Nt) ; for i\in (indI(2)+1,indI(3)], Y(i)<0 (deriv. X''(t1))
|
||||
! (indI(3)=Nt+1); for i\in (indI(3)+1,indI(4)], Y(i)>0 (deriv. X''(tn))
|
||||
! (indI(4)=Nt+2); for i\in (indI(4)+1,indI(5)], Y(i)<0 (deriv. X'(ts))
|
||||
!
|
||||
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
|
||||
!
|
||||
!Revised pab 22.04.2000
|
||||
! - added mean separated min/max + (Tdm, TMd) period distributions
|
||||
! - added scis
|
||||
|
||||
|
||||
C ***** The bound 'infinity' is set to 10*sigma *****
|
||||
XdInf = 10.d0*SQRT(R4(1))
|
||||
XtInf = 10.d0*SQRT(-R2(1))
|
||||
|
||||
Nc = 4
|
||||
NI=4; Nd=2;
|
||||
Mb=1 ;
|
||||
Nj = 0
|
||||
indI(1) = 0
|
||||
Nstart=MAX(2,Nstart)
|
||||
|
||||
isOdd = MOD(Nx1,2)
|
||||
IF (def.LE.1) THEN ! just plain Mm
|
||||
Nx = Nx1*(Nx1-1)/2
|
||||
IJ = (Nx1+isOdd)/2
|
||||
IF (H(1)+H(Nx1).EQ.0.AND.
|
||||
& (H(IJ).EQ.0.OR.H(IJ)+H(IJ+1).EQ.0) ) THEN
|
||||
SYMMETRY=.FALSE.
|
||||
PRINT *,' Integration region symmetric'
|
||||
! May save Nx1-isOdd integrations in each time step
|
||||
! This is not implemented yet.
|
||||
!Nx = Nx1*(Nx1-1)/2-Nx1+isOdd
|
||||
ENDIF
|
||||
|
||||
CC = TWOPI*SQRT(-R2(1)/R4(1)) ! normalizing constant = 1/ expected number of zero-up-crossings of X'
|
||||
|
||||
ELSE ! level u separated Mm
|
||||
Nx = (Nx1-1)*(Nx1-1)
|
||||
IF ( ABS(u).LE.1D-8.AND.H(1)+H(Nx1+1).EQ.0.AND.
|
||||
& (H(Nx1)+H(2*Nx1).EQ.0) ) THEN
|
||||
SYMMETRY=.FALSE.
|
||||
PRINT *,' Integration region symmetric'
|
||||
! Not implemented for DEF <= 3
|
||||
!IF (DEF.LE.3) Nx = (Nx1-1)*(Nx1-2)/2
|
||||
ENDIF
|
||||
|
||||
IF (DEF.GT.3) THEN
|
||||
Nstart = MAX(Nstart,3)
|
||||
Nc = 5
|
||||
NI=5; Nd=3;
|
||||
ENDIF
|
||||
CC = TWOPI*SQRT(-R0(1)/R2(1))*exp(0.5D0*u*u/R0(1)) ! normalizing constant= 1/ expected number of u-up-crossings of X
|
||||
ENDIF
|
||||
|
||||
!print *,'def',def
|
||||
IF (Nx.GT.1) THEN
|
||||
IF ((DEF.EQ.0.OR.DEF.EQ.2)) THEN ! (M,m) or (M,m)v distribution wanted
|
||||
allocate(ansr(Nx1,Nx1,1),stat=status)
|
||||
ELSE ! (M,m,TMm), (M,m,TMm)v (M,m,TMd)v or (M,M,Tdm)v distributions wanted
|
||||
allocate(ansr(Nx1,Nx1,Ntime),stat=status)
|
||||
ENDIF
|
||||
ELSEIF (DEF.GT.3) THEN ! Conditional distribution for (TMd,TMm)v or (Tdm,TMm)v given (M,m) wanted
|
||||
allocate(ansr(1,Ntime,Ntime),stat=status)
|
||||
ELSE ! Conditional distribution for (TMm) or (TMm)v given (M,m) wanted
|
||||
allocate(ansr(1,1,Ntime),stat=status)
|
||||
ENDIF
|
||||
if (status.ne.0) print *,'can not allocate ansr'
|
||||
allocate(BIG(Ntime+Nc+1,Ntime+Nc+1),stat=status)
|
||||
if (status.ne.0) print *,'can not allocate BIG'
|
||||
allocate(ex(1:Ntime+Nc+1),stat=status)
|
||||
if (status.ne.0) print *,'can not allocate ex'
|
||||
allocate(fxind(Nx),xc(Nc,Nx))
|
||||
|
||||
|
||||
! Initialization
|
||||
!~~~~~~~~~~~~~~~~~
|
||||
|
||||
BIG = 0.d0
|
||||
ex = 0.d0
|
||||
ansr = 0.d0
|
||||
a_up = 0.d0
|
||||
a_lo = 0.d0
|
||||
|
||||
xc(:,:) = 0.d0
|
||||
!xc(:,1:Nx) = 0.d0
|
||||
!xc(2,1:Nx) = 0.d0
|
||||
|
||||
a_lo(1,1) = -Xtinf
|
||||
a_lo(1,2) = -XdInf
|
||||
a_up(1,3) = +XdInf
|
||||
a_lo(1,4) = -Xtinf
|
||||
ij = 0
|
||||
IF (DEF.LE.1) THEN ! Max2min and period/wavelength
|
||||
do I=2,Nx1
|
||||
J = IJ+I-1
|
||||
xc(3,IJ+1:J) = h(I)
|
||||
xc(4,IJ+1:J) = h(1:I-1)
|
||||
IJ = J
|
||||
enddo
|
||||
ELSE
|
||||
! Level u separated Max2min
|
||||
xc(Nc,:) = u
|
||||
! H(1) = H(Nx1+1)= u => start do loop at I=2 since by definition we must have: minimum<u-level<Maximum
|
||||
do i=2,Nx1
|
||||
J = IJ+Nx1-1
|
||||
xc(3,IJ+1:J) = h(i) ! Max > u
|
||||
xc(4,IJ+1:J) = h(Nx1+2:2*Nx1) ! Min < u
|
||||
IJ = J
|
||||
enddo
|
||||
|
||||
!CALL ECHO(transpose(xc(3:5,:)))
|
||||
if (DEF.GT.3) GOTO 200
|
||||
ENDIF
|
||||
do Ntd = Nstart,Ntime
|
||||
!Ntd=tn
|
||||
Ntdc = Ntd+Nc
|
||||
Nt = Ntd-Nd;
|
||||
indI(2) = Nt;
|
||||
indI(3) = Nt+1;
|
||||
indI(4) = Ntd;
|
||||
CALL COV_INPUT(BIG(1:Ntdc,1:Ntdc),Ntd,0,R0,R1,R2,R3,R4) ! positive wave period
|
||||
CALL RINDD(fxind,Big(1:Ntdc,1:Ntdc),ex,xc,indI,a_lo,a_up)
|
||||
IF (Nx.LT.2) THEN
|
||||
! Density of TMm given the Max and the Min. Note that the density is not scaled to unity
|
||||
ansr(1,1,Ntd) = fxind(1)*CC
|
||||
GOTO 100
|
||||
ENDIF
|
||||
IJ = 0
|
||||
SELECT CASE (DEF)
|
||||
CASE(:0)
|
||||
! joint density of (M,m)
|
||||
!~~~~~~~~~~~~~~~~~~~~~~~~
|
||||
do i = 2, Nx1
|
||||
J = IJ+i-1
|
||||
ansr(1:i-1,i,1) = ansr(1:i-1,i,1)+fxind(ij+1:J)*CC*dt
|
||||
IJ=J
|
||||
enddo
|
||||
CASE (1)
|
||||
! joint density of (M,m,TMm)
|
||||
do i = 2, Nx1
|
||||
J = IJ+i-1
|
||||
ansr(1:i-1,i,Ntd) = fxind(ij+1:J)*CC
|
||||
IJ = J
|
||||
enddo
|
||||
CASE (2)
|
||||
! joint density of level v separated (M,m)v
|
||||
do i = 2,Nx1
|
||||
J = IJ+Nx1-1
|
||||
ansr(2:Nx1,i,1) = ansr(2:Nx1,i,1)+fxind(ij+1:J)*CC*dt
|
||||
IJ = J
|
||||
enddo
|
||||
CASE (3:)
|
||||
! joint density of level v separated (M,m,TMm)v
|
||||
do i = 2,Nx1
|
||||
J = IJ+Nx1-1
|
||||
ansr(2:Nx1,i,Ntd) = ansr(2:Nx1,i,Ntd)+fxind(ij+1:J)*CC
|
||||
IJ = J
|
||||
enddo
|
||||
END SELECT
|
||||
|
||||
100 if (ALLOCATED(COV)) then
|
||||
write(11,*) COV(:) ! save coefficient of variation
|
||||
endif
|
||||
print *,'Ready: ',Ntd,' of ',Ntime
|
||||
enddo
|
||||
|
||||
goto 800
|
||||
|
||||
200 do tn = Nstart,Ntime
|
||||
Ntd = tn+1
|
||||
Ntdc = Ntd + Nc
|
||||
Nt = Ntd - Nd;
|
||||
indI(2) = Nt;
|
||||
indI(3) = Nt + 1;
|
||||
indI(4) = Nt + 2;
|
||||
indI(5) = Ntd;
|
||||
!CALL COV_INPUT2(BIG(1:Ntdc,1:Ntdc),tn,-2,R0,R1,R2,R3,R4) ! positive wave period
|
||||
IF (SYMMETRY) GOTO 300
|
||||
|
||||
do ts = 2,tn-1
|
||||
CALL COV_INPUT(BIG(1:Ntdc,1:Ntdc),tn,ts,R0,R1,R2,R3,R4) ! positive wave period
|
||||
!print *,'Big='
|
||||
!CALL ECHO(BIG(1:Ntdc,1:MIN(Ntdc,10)))
|
||||
CALL RINDD(fxind,Big(1:Ntdc,1:Ntdc),ex,xc,indI,a_lo,a_up)
|
||||
|
||||
SELECT CASE (def)
|
||||
CASE (:4)
|
||||
IF (Nx.EQ.1) THEN
|
||||
! Joint density (TMd,TMm) given the Max and the min. Note the density is not scaled to unity
|
||||
ansr(1,ts,tn) = fxind(1)*CC
|
||||
|
||||
ELSE
|
||||
! 4, gives level u separated Max2min and wave period from Max to the crossing of level u (M,m,TMd).
|
||||
ij = 0
|
||||
do i = 2,Nx1
|
||||
J = IJ+Nx1-1
|
||||
ansr(2:Nx1,i,ts) = ansr(2:Nx1,i,ts)+
|
||||
& fxind(ij+1:J)*CC*dt
|
||||
IJ = J
|
||||
enddo
|
||||
ENDIF
|
||||
CASE (5:)
|
||||
IF (Nx.EQ.1) THEN
|
||||
! Joint density (Tdm,TMm) given the Max and the min. Note the density is not scaled to unity
|
||||
ansr(1,tn-ts+1,tn) = fxind(1)*CC
|
||||
ELSE
|
||||
|
||||
! 5, gives level u separated Max2min and wave period from the crossing of level u to the min (M,m,Tdm).
|
||||
ij = 0
|
||||
do i = 2,Nx1
|
||||
J = IJ+Nx1-1
|
||||
ansr(2:Nx1,i,tn-ts+1)=ansr(2:Nx1,i,tn-ts+1)+
|
||||
& fxind(ij+1:J)*CC*dt
|
||||
IJ = J
|
||||
enddo
|
||||
ENDIF
|
||||
END SELECT
|
||||
if (ALLOCATED(COV)) then
|
||||
write(11,*) COV(:) ! save coefficient of variation
|
||||
endif
|
||||
enddo
|
||||
GOTO 400
|
||||
300 do ts = 2,FLOOR(DBLE(Ntd)/2.d0) ! Using the symmetry since U = 0 and the transformation is linear
|
||||
CALL COV_INPUT(BIG(1:Ntdc,1:Ntdc),tn,ts,R0,R1,R2,R3,R4) ! positive wave period
|
||||
!print *,'Big='
|
||||
!CALL ECHO(BIG(1:Ntdc,1:Ntdc))
|
||||
CALL RINDD(fxind,Big(1:Ntdc,1:Ntdc),ex,xc,indI,a_lo,a_up)
|
||||
IF (Nx.EQ.1) THEN
|
||||
! Joint density of (TMd,TMm),(Tdm,TMm) given the max and the min. Note that the density is not scaled to unity
|
||||
ansr(1,ts,tn) = fxind(1)*CC
|
||||
IF (ts.LT.tn-ts+1) THEN
|
||||
ansr(1,tn-ts+1,tn) = fxind(1)*CC
|
||||
ENDIF
|
||||
GOTO 350
|
||||
ENDIF
|
||||
IJ = 0
|
||||
SELECT CASE (def)
|
||||
CASE (:4)
|
||||
|
||||
! 4, gives level u separated Max2min and wave period from Max to the crossing of level u (M,m,TMd).
|
||||
do i = 2,Nx1
|
||||
j = ij+Nx1-1
|
||||
ansr(2:Nx1,i,ts) = ansr(2:Nx1,i,ts)+
|
||||
& fxind(ij+1:J)*CC*dt
|
||||
IF (ts.LT.tn-ts+1) THEN
|
||||
ansr(i,2:Nx1,tn-ts+1) =
|
||||
& ansr(i,2:Nx1,tn-ts+1)+fxind(ij+1:J)*CC*dt ! exploiting the symmetry
|
||||
ENDIF
|
||||
IJ = J
|
||||
enddo
|
||||
CASE (5:)
|
||||
! 5, gives level u separated Max2min and wave period from the crossing of level u to min (M,m,Tdm).
|
||||
do i = 2,Nx1
|
||||
J = IJ+Nx1-1
|
||||
|
||||
ansr(2:Nx1,i,tn-ts+1)=ansr(2:Nx1,i,tn-ts+1)+
|
||||
& fxind(ij+1:J)*CC*dt
|
||||
IF (ts.LT.tn-ts+1) THEN
|
||||
ansr(i,2:Nx1,ts) = ansr(i,2:Nx1,ts)+
|
||||
& fxind(ij+1:J)*CC*dt ! exploiting the symmetry
|
||||
ENDIF
|
||||
IJ = J
|
||||
enddo
|
||||
END SELECT
|
||||
350 enddo
|
||||
400 print *,'Ready: ',tn,' of ',Ntime
|
||||
enddo
|
||||
|
||||
|
||||
|
||||
|
||||
800 open (unit=11, file='dens.out', STATUS='unknown')
|
||||
!print *,'ans, IJ,def', shape(ansr),IJ,DEF
|
||||
if (Nx.GT.1) THEN
|
||||
ij = 1
|
||||
IF (DEF.GT.2.OR.DEF.EQ.1) IJ = Ntime
|
||||
!print *,'ans, IJ,def', size(ansr),IJ,DEF
|
||||
do ts = 1,ij
|
||||
do j=1,Nx1
|
||||
do i=1,Nx1
|
||||
write(11,*) ansr(i,j,ts)
|
||||
enddo
|
||||
enddo
|
||||
enddo
|
||||
ELSE
|
||||
ij = 1
|
||||
IF (DEF.GT.3) IJ = Ntime
|
||||
!print *,'ans, IJ,def', size(ansr),IJ,DEF
|
||||
do ts = 1,Ntime
|
||||
do j = 1,ij
|
||||
write(11,*) ansr(1,j,ts)
|
||||
enddo
|
||||
enddo
|
||||
ENDIF
|
||||
close(11)
|
||||
900 continue
|
||||
deallocate(BIG)
|
||||
deallocate(ex)
|
||||
deallocate(fxind)
|
||||
deallocate(ansr)
|
||||
deallocate(xc)
|
||||
deallocate(R0)
|
||||
deallocate(R1)
|
||||
deallocate(R2)
|
||||
deallocate(R3)
|
||||
deallocate(R4)
|
||||
deallocate(h)
|
||||
|
||||
if (allocated(COV) ) then
|
||||
deallocate(COV)
|
||||
endif
|
||||
stop
|
||||
!return
|
||||
|
||||
CONTAINS
|
||||
|
||||
|
||||
|
||||
SUBROUTINE INIT_LEVELS
|
||||
& (Ntime,Nstart,NIT,speed,SCIS,SEED1,Nx,dT,u,def)
|
||||
IMPLICIT NONE
|
||||
integer, intent(out):: Ntime,Nstart,NIT,speed,Nx,DEF,SCIS,SEED1
|
||||
double precision ,intent(out) :: dT,U
|
||||
|
||||
|
||||
OPEN(UNIT=14,FILE='reflev.in',STATUS= 'UNKNOWN')
|
||||
READ (14,*) Ntime
|
||||
READ (14,*) Nstart
|
||||
READ (14,*) NIT
|
||||
READ (14,*) speed
|
||||
READ (14,*) SCIS
|
||||
READ (14,*) seed1
|
||||
READ (14,*) Nx
|
||||
READ (14,*) dT
|
||||
READ (14,*) U
|
||||
READ (14,*) DEF
|
||||
|
||||
if (Ntime.lt.2) then
|
||||
print *,'The number of wavelength points is too small, stop'
|
||||
stop
|
||||
end if
|
||||
CLOSE(UNIT=14)
|
||||
|
||||
RETURN
|
||||
END SUBROUTINE INIT_LEVELS
|
||||
|
||||
C******************************************************
|
||||
SUBROUTINE INIT_AMPLITUDES(h,Nx)
|
||||
IMPLICIT NONE
|
||||
double precision, dimension(:), intent(out) :: h
|
||||
integer, intent(in) :: Nx
|
||||
integer :: ix
|
||||
|
||||
|
||||
OPEN(UNIT=4,FILE='h.in',STATUS= 'UNKNOWN')
|
||||
|
||||
C
|
||||
C Reading in amplitudes
|
||||
C
|
||||
do ix=1,Nx
|
||||
READ (4,*) H(ix)
|
||||
enddo
|
||||
CLOSE(UNIT=4)
|
||||
|
||||
RETURN
|
||||
END SUBROUTINE INIT_AMPLITUDES
|
||||
|
||||
C**************************************************
|
||||
|
||||
C***********************************************************************
|
||||
C***********************************************************************
|
||||
|
||||
SUBROUTINE INIT_COVARIANCES(Ntime,R0,R1,R2,R3,R4)
|
||||
IMPLICIT NONE
|
||||
double precision, dimension(:),intent(out) :: R0,R1,R2
|
||||
double precision, dimension(:),intent(out) :: R3,R4
|
||||
integer,intent(in) :: Ntime
|
||||
integer :: i
|
||||
open (unit=1, file='Cd0.in',STATUS='unknown')
|
||||
open (unit=2, file='Cd1.in',STATUS='unknown')
|
||||
open (unit=3, file='Cd2.in',STATUS='unknown')
|
||||
open (unit=4, file='Cd3.in',STATUS='unknown')
|
||||
open (unit=5, file='Cd4.in',STATUS='unknown')
|
||||
|
||||
do i=1,Ntime
|
||||
read(1,*) R0(i)
|
||||
read(2,*) R1(i)
|
||||
read(3,*) R2(i)
|
||||
read(4,*) R3(i)
|
||||
read(5,*) R4(i)
|
||||
enddo
|
||||
close(1)
|
||||
close(2)
|
||||
close(3)
|
||||
close(3)
|
||||
close(5)
|
||||
return
|
||||
END SUBROUTINE INIT_COVARIANCES
|
||||
|
||||
C**********************************************************************
|
||||
|
||||
SUBROUTINE COV_INPUT2(BIG,tn,ts,R0,R1,R2,R3,R4)
|
||||
IMPLICIT NONE
|
||||
double precision, dimension(:,:),intent(inout) :: BIG
|
||||
double precision, dimension(:),intent(in) :: R0,R1,R2
|
||||
double precision, dimension(:),intent(in) :: R3,R4
|
||||
integer ,intent(in) :: tn,ts
|
||||
integer :: i,j,N,shft
|
||||
! the order of the variables in the covariance matrix
|
||||
! are organized as follows:
|
||||
! for ts <= 1:
|
||||
! X'(t2)..X'(ts),...,X'(tn-1) X''(t1),X''(tn) X'(t1),X'(tn),X(t1),X(tn)
|
||||
! = [ Xt | Xd | Xc ]
|
||||
!
|
||||
! for ts > =2:
|
||||
! X'(t2)..X'(ts),...,X'(tn-1) X''(t1),X''(tn) X'(t1),X'(tn),X(t1),X(tn) X(ts)
|
||||
! = [ Xt | Xd | Xc ]
|
||||
!
|
||||
! where
|
||||
!
|
||||
! Xt= time points in the indicator function
|
||||
! Xd= derivatives
|
||||
! Xc=variables to condition on
|
||||
|
||||
! Computations of all covariances follows simple rules: Cov(X(t),X(s))=r(t,s),
|
||||
! then Cov(X'(t),X(s))=dr(t,s)/dt. Now for stationary X(t) we have
|
||||
! a function r(tau) such that Cov(X(t),X(s))=r(s-t) (or r(t-s) will give the same result).
|
||||
!
|
||||
! Consequently Cov(X'(t),X(s)) = -r'(s-t) = -sign(s-t)*r'(|s-t|)
|
||||
! Cov(X'(t),X'(s)) = -r''(s-t) = -r''(|s-t|)
|
||||
! Cov(X''(t),X'(s)) = r'''(s-t) = sign(s-t)*r'''(|s-t|)
|
||||
! Cov(X''(t),X(s)) = r''(s-t) = r''(|s-t|)
|
||||
! Cov(X''(t),X''(s)) = r''''(s-t) = r''''(|s-t|)
|
||||
|
||||
if (ts.GT.1) THEN
|
||||
! Assumption: a previous call to covinput has been made
|
||||
! need only to update the last row and column of big:
|
||||
N=tn+5
|
||||
!Cov(Xt,Xc)
|
||||
do i=1,tn-2
|
||||
j=abs(i+1-ts)
|
||||
BIG(i,N) = -sign(R1(j+1),R1(j+1)*dble(ts-i-1)) !cov(X'(ti+1),X(ts))
|
||||
enddo
|
||||
!Cov(Xc)
|
||||
BIG(N ,N) = R0(1) ! cov(X(ts),X(ts))
|
||||
BIG(tn+3 ,N) = R0(ts) ! cov(X(t1),X(ts))
|
||||
BIG(tn+4 ,N) = R0(tn-ts+1) ! cov(X(tn),X(ts))
|
||||
BIG(tn+1 ,N) = -R1(ts) ! cov(X'(t1),X(ts))
|
||||
BIG(tn+2 ,N) = R1(tn-ts+1) ! cov(X'(tn),X(ts))
|
||||
!Cov(Xd,Xc)
|
||||
BIG(tn-1 ,N) = R2(ts) !cov(X''(t1),X(ts))
|
||||
BIG(tn ,N) = R2(tn-ts+1) !cov(X''(tn),X(ts))
|
||||
|
||||
! make lower triangular part equal to upper
|
||||
do j=1,N-1
|
||||
BIG(N,j) = BIG(j,N)
|
||||
enddo
|
||||
return
|
||||
endif
|
||||
IF (ts.LT.0) THEN
|
||||
shft = 1
|
||||
N=tn+5;
|
||||
ELSE
|
||||
shft = 0
|
||||
N=tn+4;
|
||||
ENDIF
|
||||
|
||||
|
||||
do i=1,tn-2
|
||||
!cov(Xt)
|
||||
do j=i,tn-2
|
||||
BIG(i,j) = -R2(j-i+1) ! cov(X'(ti+1),X'(tj+1))
|
||||
enddo
|
||||
!cov(Xt,Xc)
|
||||
BIG(i ,tn+3) = R1(i+1) !cov(X'(ti+1),X(t1))
|
||||
BIG(tn-1-i ,tn+4) = -R1(i+1) !cov(X'(ti+1),X(tn))
|
||||
BIG(i ,tn+1) = -R2(i+1) !cov(X'(ti+1),X'(t1))
|
||||
BIG(tn-1-i ,tn+2) = -R2(i+1) !cov(X'(ti+1),X'(tn))
|
||||
!Cov(Xt,Xd)
|
||||
BIG(i,tn-1) = R3(i+1) !cov(X'(ti+1),X''(t1))
|
||||
BIG(tn-1-i,tn) =-R3(i+1) !cov(X'(ti+1),X''(tn))
|
||||
enddo
|
||||
|
||||
!cov(Xd)
|
||||
BIG(tn-1 ,tn-1 ) = R4(1)
|
||||
BIG(tn-1 ,tn ) = R4(tn) !cov(X''(t1),X''(tn))
|
||||
BIG(tn ,tn ) = R4(1)
|
||||
|
||||
!cov(Xc)
|
||||
BIG(tn+3 ,tn+3) = R0(1) ! cov(X(t1),X(t1))
|
||||
BIG(tn+3 ,tn+4) = R0(tn) ! cov(X(t1),X(tn))
|
||||
BIG(tn+1 ,tn+3) = 0.d0 ! cov(X(t1),X'(t1))
|
||||
BIG(tn+2 ,tn+3) = R1(tn) ! cov(X(t1),X'(tn))
|
||||
BIG(tn+4 ,tn+4) = R0(1) ! cov(X(tn),X(tn))
|
||||
BIG(tn+1 ,tn+4) =-R1(tn) ! cov(X(tn),X'(t1))
|
||||
BIG(tn+2 ,tn+4) = 0.d0 ! cov(X(tn),X'(tn))
|
||||
BIG(tn+1 ,tn+1) =-R2(1) ! cov(X'(t1),X'(t1))
|
||||
BIG(tn+1 ,tn+2) =-R2(tn) ! cov(X'(t1),X'(tn))
|
||||
BIG(tn+2 ,tn+2) =-R2(1) ! cov(X'(tn),X'(tn))
|
||||
!Xc=X(t1),X(tn),X'(t1),X'(tn)
|
||||
!Xd=X''(t1),X''(tn)
|
||||
!cov(Xd,Xc)
|
||||
BIG(tn-1 ,tn+3) = R2(1) !cov(X''(t1),X(t1))
|
||||
BIG(tn-1 ,tn+4) = R2(tn) !cov(X''(t1),X(tn))
|
||||
BIG(tn-1 ,tn+1) = 0.d0 !cov(X''(t1),X'(t1))
|
||||
BIG(tn-1 ,tn+2) = R3(tn) !cov(X''(t1),X'(tn))
|
||||
BIG(tn ,tn+3) = R2(tn) !cov(X''(tn),X(t1))
|
||||
BIG(tn ,tn+4) = R2(1) !cov(X''(tn),X(tn))
|
||||
BIG(tn ,tn+1) =-R3(tn) !cov(X''(tn),X'(t1))
|
||||
BIG(tn ,tn+2) = 0.d0 !cov(X''(tn),X'(tn))
|
||||
|
||||
|
||||
! make lower triangular part equal to upper
|
||||
do j=1,N-1
|
||||
do i=j+1,N
|
||||
BIG(i,j) = BIG(j,i)
|
||||
enddo
|
||||
enddo
|
||||
RETURN
|
||||
END SUBROUTINE COV_INPUT2
|
||||
|
||||
SUBROUTINE COV_INPUT(BIG,tn,ts,R0,R1,R2,R3,R4)
|
||||
IMPLICIT NONE
|
||||
double precision, dimension(:,:),intent(inout) :: BIG
|
||||
double precision, dimension(:),intent(in) :: R0,R1,R2
|
||||
double precision, dimension(:),intent(in) :: R3,R4
|
||||
integer ,intent(in) :: tn,ts
|
||||
integer :: i,j,N,shft, tnold = 0
|
||||
! the order of the variables in the covariance matrix
|
||||
! are organized as follows:
|
||||
! for ts <= 1:
|
||||
! X'(t2)..X'(ts),...,X'(tn-1) X''(t1),X''(tn) X'(t1),X'(tn),X(t1),X(tn)
|
||||
! = [ Xt | Xd | Xc ]
|
||||
!
|
||||
! for ts > =2:
|
||||
! X'(t2)..X'(ts),...,X'(tn-1) X''(t1),X''(tn) X'(ts) X'(t1),X'(tn),X(t1),X(tn) X(ts)
|
||||
! = [ Xt | Xd | Xc ]
|
||||
!
|
||||
! where
|
||||
!
|
||||
! Xt= time points in the indicator function
|
||||
! Xd= derivatives
|
||||
! Xc=variables to condition on
|
||||
|
||||
! Computations of all covariances follows simple rules: Cov(X(t),X(s)) = r(t,s),
|
||||
! then Cov(X'(t),X(s))=dr(t,s)/dt. Now for stationary X(t) we have
|
||||
! a function r(tau) such that Cov(X(t),X(s))=r(s-t) (or r(t-s) will give the same result).
|
||||
!
|
||||
! Consequently Cov(X'(t),X(s)) = -r'(s-t) = -sign(s-t)*r'(|s-t|)
|
||||
! Cov(X'(t),X'(s)) = -r''(s-t) = -r''(|s-t|)
|
||||
! Cov(X''(t),X'(s)) = r'''(s-t) = sign(s-t)*r'''(|s-t|)
|
||||
! Cov(X''(t),X(s)) = r''(s-t) = r''(|s-t|)
|
||||
! Cov(X''(t),X''(s)) = r''''(s-t) = r''''(|s-t|)
|
||||
SAVE tnold
|
||||
|
||||
if (ts.GT.1) THEN
|
||||
shft = 1
|
||||
N=tn+5+shft
|
||||
!Cov(Xt,Xc)
|
||||
do i=1,tn-2
|
||||
j=abs(i+1-ts)
|
||||
BIG(i,N) = -sign(R1(j+1),R1(j+1)*dble(ts-i-1)) !cov(X'(ti+1),X(ts))
|
||||
enddo
|
||||
!Cov(Xc)
|
||||
BIG(N ,N) = R0(1) ! cov(X(ts),X(ts))
|
||||
BIG(tn+shft+3 ,N) = R0(ts) ! cov(X(t1),X(ts))
|
||||
BIG(tn+shft+4 ,N) = R0(tn-ts+1) ! cov(X(tn),X(ts))
|
||||
BIG(tn+shft+1 ,N) = -R1(ts) ! cov(X'(t1),X(ts))
|
||||
BIG(tn+shft+2 ,N) = R1(tn-ts+1) ! cov(X'(tn),X(ts))
|
||||
!Cov(Xd,Xc)
|
||||
BIG(tn-1 ,N) = R2(ts) !cov(X''(t1),X(ts))
|
||||
BIG(tn ,N) = R2(tn-ts+1) !cov(X''(tn),X(ts))
|
||||
|
||||
!ADD a level u crossing at ts
|
||||
|
||||
!Cov(Xt,Xd)
|
||||
do i = 1,tn-2
|
||||
j = abs(i+1-ts)
|
||||
BIG(i,tn+shft) = -R2(j+1) !cov(X'(ti+1),X'(ts))
|
||||
enddo
|
||||
!Cov(Xd)
|
||||
BIG(tn+shft,tn+shft) = -R2(1) !cov(X'(ts),X'(ts))
|
||||
BIG(tn-1 ,tn+shft) = R3(ts) !cov(X''(t1),X'(ts))
|
||||
BIG(tn ,tn+shft) = -R3(tn-ts+1) !cov(X''(tn),X'(ts))
|
||||
|
||||
!Cov(Xd,Xc)
|
||||
BIG(tn+shft ,N ) = 0.d0 !cov(X'(ts),X(ts))
|
||||
BIG(tn+shft,tn+shft+3) = R1(ts) ! cov(X'(ts),X(t1))
|
||||
BIG(tn+shft,tn+shft+4) = -R1(tn-ts+1) ! cov(X'(ts),X(tn))
|
||||
BIG(tn+shft,tn+shft+1) = -R2(ts) ! cov(X'(ts),X'(t1))
|
||||
BIG(tn+shft,tn+shft+2) = -R2(tn-ts+1) ! cov(X'(ts),X'(tn))
|
||||
|
||||
|
||||
|
||||
IF (tnold.EQ.tn) THEN ! A previous call to covinput with tn==tnold has been made
|
||||
! need only to update row and column N and tn+1 of big:
|
||||
! make lower triangular part equal to upper and then return
|
||||
do j=1,tn+shft
|
||||
BIG(N,j) = BIG(j,N)
|
||||
BIG(tn+shft,j) = BIG(j,tn+shft)
|
||||
enddo
|
||||
do j=tn+shft+1,N-1
|
||||
BIG(N,j) = BIG(j,N)
|
||||
BIG(j,tn+shft) = BIG(tn+shft,j)
|
||||
enddo
|
||||
return
|
||||
ENDIF
|
||||
tnold = tn
|
||||
ELSE
|
||||
N = tn+4
|
||||
shft = 0
|
||||
endif
|
||||
|
||||
|
||||
do i=1,tn-2
|
||||
!cov(Xt)
|
||||
do j=i,tn-2
|
||||
BIG(i,j) = -R2(j-i+1) ! cov(X'(ti+1),X'(tj+1))
|
||||
enddo
|
||||
!cov(Xt,Xc)
|
||||
BIG(i ,tn+shft+3) = R1(i+1) !cov(X'(ti+1),X(t1))
|
||||
BIG(tn-1-i ,tn+shft+4) = -R1(i+1) !cov(X'(ti+1),X(tn))
|
||||
BIG(i ,tn+shft+1) = -R2(i+1) !cov(X'(ti+1),X'(t1))
|
||||
BIG(tn-1-i ,tn+shft+2) = -R2(i+1) !cov(X'(ti+1),X'(tn))
|
||||
!Cov(Xt,Xd)
|
||||
BIG(i,tn-1) = R3(i+1) !cov(X'(ti+1),X''(t1))
|
||||
BIG(tn-1-i,tn) =-R3(i+1) !cov(X'(ti+1),X''(tn))
|
||||
enddo
|
||||
|
||||
!cov(Xd)
|
||||
BIG(tn-1 ,tn-1 ) = R4(1)
|
||||
BIG(tn-1 ,tn ) = R4(tn) !cov(X''(t1),X''(tn))
|
||||
BIG(tn ,tn ) = R4(1)
|
||||
|
||||
!cov(Xc)
|
||||
BIG(tn+shft+3 ,tn+shft+3) = R0(1) ! cov(X(t1),X(t1))
|
||||
BIG(tn+shft+3 ,tn+shft+4) = R0(tn) ! cov(X(t1),X(tn))
|
||||
BIG(tn+shft+1 ,tn+shft+3) = 0.d0 ! cov(X(t1),X'(t1))
|
||||
BIG(tn+shft+2 ,tn+shft+3) = R1(tn) ! cov(X(t1),X'(tn))
|
||||
BIG(tn+shft+4 ,tn+shft+4) = R0(1) ! cov(X(tn),X(tn))
|
||||
BIG(tn+shft+1 ,tn+shft+4) =-R1(tn) ! cov(X(tn),X'(t1))
|
||||
BIG(tn+shft+2 ,tn+shft+4) = 0.d0 ! cov(X(tn),X'(tn))
|
||||
BIG(tn+shft+1 ,tn+shft+1) =-R2(1) ! cov(X'(t1),X'(t1))
|
||||
BIG(tn+shft+1 ,tn+shft+2) =-R2(tn) ! cov(X'(t1),X'(tn))
|
||||
BIG(tn+shft+2 ,tn+shft+2) =-R2(1) ! cov(X'(tn),X'(tn))
|
||||
!Xc=X(t1),X(tn),X'(t1),X'(tn)
|
||||
!Xd=X''(t1),X''(tn)
|
||||
!cov(Xd,Xc)
|
||||
BIG(tn-1 ,tn+shft+3) = R2(1) !cov(X''(t1),X(t1))
|
||||
BIG(tn-1 ,tn+shft+4) = R2(tn) !cov(X''(t1),X(tn))
|
||||
BIG(tn-1 ,tn+shft+1) = 0.d0 !cov(X''(t1),X'(t1))
|
||||
BIG(tn-1 ,tn+shft+2) = R3(tn) !cov(X''(t1),X'(tn))
|
||||
BIG(tn ,tn+shft+3) = R2(tn) !cov(X''(tn),X(t1))
|
||||
BIG(tn ,tn+shft+4) = R2(1) !cov(X''(tn),X(tn))
|
||||
BIG(tn ,tn+shft+1) =-R3(tn) !cov(X''(tn),X'(t1))
|
||||
BIG(tn ,tn+shft+2) = 0.d0 !cov(X''(tn),X'(tn))
|
||||
|
||||
|
||||
! make lower triangular part equal to upper
|
||||
do j=1,N-1
|
||||
do i=j+1,N
|
||||
BIG(i,j) = BIG(j,i)
|
||||
enddo
|
||||
enddo
|
||||
RETURN
|
||||
END SUBROUTINE COV_INPUT
|
||||
END PROGRAM sp2mmt
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
@ -1,498 +0,0 @@
|
||||
PROGRAM sp2tccpdf
|
||||
C***********************************************************************
|
||||
C This program computes: *
|
||||
C *
|
||||
C density of T= T_1+T_2 in a gaussian process i.e. *
|
||||
C *
|
||||
C wavelengthes for crests <h1 and troughs >h2 *
|
||||
C *
|
||||
C Sylvie and Igor 7 dec. 1999 *
|
||||
C***********************************************************************
|
||||
use GLOBALDATA, only : Nt,Nj,Nd,Nc,Ntd,Ntdc,NI,Mb,
|
||||
& NIT,Nx,TWOPI,XSPLT,SCIS,NSIMmax,COV
|
||||
use rind
|
||||
IMPLICIT NONE
|
||||
double precision, dimension(:,:),allocatable :: BIG
|
||||
double precision, dimension(:,:),allocatable :: ansr
|
||||
double precision, dimension(: ),allocatable :: ex,CY1,CY2
|
||||
double precision, dimension(:,:),allocatable :: xc
|
||||
double precision, dimension(: ),allocatable :: fxind,h1,h2
|
||||
double precision, dimension(: ),allocatable :: hh1,hh2
|
||||
double precision, dimension(: ),allocatable :: R0,R1,R2
|
||||
double precision ::CC,U,XddInf,XdInf,XtInf
|
||||
double precision, dimension(:,:),allocatable :: a_up,a_lo
|
||||
integer , dimension(: ),allocatable :: seed
|
||||
integer ,dimension(7) :: indI
|
||||
integer :: Ntime,tn,ts,speed,ph,seed1,seed_size,Nx1,Nx2,N0
|
||||
integer :: icy,icy2
|
||||
double precision :: ds,dT ! lag spacing for covariances
|
||||
! DIGITAL:
|
||||
! f90 -g2 -C -automatic -o ~/WAT/V4/sp2tthpdf1.exe rind49.f sp2tthpdf1.f
|
||||
! SOLARIS:
|
||||
!f90 -g -O -w3 -Bdynamic -fixed -o ../sp2tthpdf.exe rind49.f sp2tthpdf1.f
|
||||
|
||||
!print *,'enter sp2thpdf'
|
||||
CALL INIT_LEVELS(U,Ntime,N0,NIT,speed,SCIS,seed1,Nx1,Nx2,dT)
|
||||
|
||||
!print *,'U,Ntime,NIT,speed,SCIS,seed1,Nx,dT'
|
||||
!print *,U,Ntime,NIT,speed,SCIS,seed1,Nx,dT
|
||||
!Nx1=1
|
||||
!Nx2=1
|
||||
|
||||
Nx=Nx1*Nx1
|
||||
!print *,'NN',Nx1,Nx2,Nx
|
||||
|
||||
|
||||
!XSPLT=1.5d0
|
||||
if (SCIS.GT.0) then
|
||||
allocate(COV(1:Nx))
|
||||
call random_seed(SIZE=seed_size)
|
||||
allocate(seed(seed_size))
|
||||
call random_seed(GET=seed(1:seed_size)) ! get current seed
|
||||
seed(1)=seed1 ! change seed
|
||||
call random_seed(PUT=seed(1:seed_size))
|
||||
deallocate(seed)
|
||||
endif
|
||||
CALL INITDATA(speed)
|
||||
!print *,ntime,speed,u,NIT
|
||||
allocate(R0(1:Ntime+1))
|
||||
allocate(R1(1:Ntime+1))
|
||||
allocate(R2(1:Ntime+1))
|
||||
|
||||
allocate(h1(1:Nx1))
|
||||
allocate(h2(1:Nx2))
|
||||
CALL INIT_AMPLITUDES(h1,Nx1,h2,Nx2)
|
||||
CALL INIT_COVARIANCES(Ntime,R0,R1,R2)
|
||||
|
||||
|
||||
allocate(hh1(1:Nx))
|
||||
allocate(hh2(1:Nx))
|
||||
!h transformation
|
||||
do icy=1,Nx1
|
||||
do icy2=1,Nx2
|
||||
hh1((icy-1)*Nx2+icy2)=h1(icy);
|
||||
hh2((icy-1)*Nx2+icy2)=h2(icy2);
|
||||
enddo
|
||||
enddo
|
||||
|
||||
Nj=0
|
||||
indI(1)=0
|
||||
|
||||
C ***** The bound 'infinity' is set to 10*sigma *****
|
||||
XdInf=10.d0*SQRT(-R2(1))
|
||||
XtInf=10.d0*SQRT(R0(1))
|
||||
!h1(1)=XtInf
|
||||
!h2(1)=XtInf
|
||||
! normalizing constant
|
||||
CC=TWOPI*SQRT(-R0(1)/R2(1))*exp(u*u/(2.d0*R0(1)) )
|
||||
allocate(CY1(1:Nx))
|
||||
allocate(CY2(1:Nx))
|
||||
do icy=1,Nx
|
||||
CY1(icy)=exp(-0.5*hh1(icy)*hh1(icy)/100)/(10*sqrt(twopi))
|
||||
CY2(icy)=exp(-0.5*hh2(icy)*hh2(icy)/100)/(10*sqrt(twopi))
|
||||
enddo
|
||||
!print *,CY1
|
||||
allocate(ansr(1:Ntime,1:Nx))
|
||||
ansr=0.d0
|
||||
allocate(fxind(1:Nx))
|
||||
fxind=0.d0
|
||||
|
||||
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
|
||||
! Y={X(t2)..,X(ts),..X(tn-1)||X'(ts) X'(t1) X'(tn)||Y1 Y2 X(ts) X(t1) X(tn)} !!
|
||||
! = [Xt Xd Xc] !!
|
||||
! !!
|
||||
! Nt=tn-2, Nd=3, Nc=2+3 !!
|
||||
! !!
|
||||
! Xt= contains Nt time points in the indicator function !!
|
||||
! Xd= " Nd derivatives !!
|
||||
! Xc= " Nc variables to condition on !!
|
||||
! (Y1,Y2) dummy variables ind. of all other v. inputing h1,h2 into rindd !!
|
||||
! !!
|
||||
! There are 6 ( NI=7) regions with constant bariers: !!
|
||||
! (indI(1)=0); for i\in (indI(1),indI(2)] u<Y(i)<h1 !!
|
||||
! (indI(2)=ts-2); for i\in (indI(2),indI(2)], inf<Y(i)<inf (no restr.) !!
|
||||
! (indI(3)=ts-1); for i\in (indI(3),indI(4)], h2 <Y(i)<u !!
|
||||
! (indI(4)=Nt) ; for i\in (indI(4),indI(5)], Y(i)<0 (deriv. X'(ts)) !!
|
||||
! (indI(5)=Nt+1); for i\in (indI(5),indI(6)], Y(i)>0 (deriv. X'(t1)) !!
|
||||
! (indI(6)=Nt+2); for i\in (indI(6),indI(7)], Y(i)>0 (deriv. X'(tn)) !!
|
||||
! (indI(7)=Nt+3); NI=7. !!
|
||||
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
|
||||
|
||||
|
||||
NI=7; Nd=3
|
||||
Nc=5; Mb=3
|
||||
allocate(a_up(1:Mb,1:(NI-1)))
|
||||
allocate(a_lo(1:Mb,1:(NI-1)))
|
||||
a_up=0.d0
|
||||
a_lo=0.d0
|
||||
allocate(BIG(1:(Ntime+Nc+1),1:(Ntime+Nc+1)))
|
||||
ALLOCATE(xc(1:Nc,1:Nx))
|
||||
allocate(ex(1:(Ntime+Nc+1)))
|
||||
!print *,size(ex),Ntime
|
||||
ex=0.d0
|
||||
!print *,size(ex),ex
|
||||
xc(1,1:Nx)=hh1(1:Nx)
|
||||
xc(2,1:Nx)=hh2(1:Nx)
|
||||
xc(3,1:Nx)=u
|
||||
xc(4,1:Nx)=u
|
||||
xc(5,1:Nx)=u
|
||||
! upp- down- upp-crossings at t1,ts,tn
|
||||
|
||||
a_lo(1,1)=u
|
||||
a_up(1,2)=XtInf ! X(ts) is redundant
|
||||
a_lo(1,2)=-Xtinf
|
||||
a_up(1,3)=u
|
||||
|
||||
|
||||
a_lo(1,4)=-XdInf
|
||||
a_up(1,5)= XdInf
|
||||
a_up(1,6)= XdInf
|
||||
|
||||
a_up(2,1)=1.d0
|
||||
a_lo(3,3)=1.d0 !signe a voir!!!!!!
|
||||
! print *,a_up
|
||||
! print *,a_lo
|
||||
do tn=N0,Ntime,1
|
||||
! do tn=Ntime,Ntime,1
|
||||
Ntd=tn+1
|
||||
Nt=Ntd-Nd
|
||||
Ntdc=Ntd+Nc
|
||||
indI(4)=Nt
|
||||
indI(5)=Nt+1
|
||||
indI(6)=Nt+2
|
||||
indI(7)=Ntd
|
||||
if (SCIS.gt.0) then
|
||||
if (SCIS.EQ.2) then
|
||||
Nj=max(Nt,0)
|
||||
else
|
||||
Nj=min(max(Nt-5, 0),0)
|
||||
endif
|
||||
endif
|
||||
do ts=3,tn-2
|
||||
!print *,'ts,tn' ,ts,tn,Ntdc
|
||||
CALL COV_INPUT(Big(1:Ntdc,1:Ntdc),tn,ts,R0,R1,R2)!positive wave period
|
||||
indI(2)=ts-2
|
||||
indI(3)=ts-1
|
||||
|
||||
|
||||
CALL RINDD(fxind,Big(1:Ntdc,1:Ntdc),ex(1:Ntdc),
|
||||
& xc,indI,a_lo,a_up)
|
||||
|
||||
ds=dt
|
||||
do icy=1,Nx
|
||||
! ansr(tn,:)=ansr(tn,:)+fxind*CC*ds./(CY1.*CY2)
|
||||
ansr(tn,icy)=ansr(tn,icy)+fxind(icy)*CC*ds/(CY1(icy)*CY2(icy))
|
||||
enddo
|
||||
enddo ! ts
|
||||
print *,'Ready: ',tn,' of ',Ntime
|
||||
|
||||
enddo !tn
|
||||
!print *,'ansr',ansr
|
||||
300 open (unit=11, file='dens.out', STATUS='unknown')
|
||||
!print *, ansr
|
||||
do ts=1,Ntime
|
||||
do ph=1,Nx
|
||||
!write(11,*) ansr(ts,ph),hh1(ph),hh2(ph)
|
||||
write(11,111) ansr(ts,ph)
|
||||
|
||||
enddo
|
||||
enddo
|
||||
111 FORMAT(2x,F12.8)
|
||||
close(11)
|
||||
900 deallocate(big)
|
||||
deallocate(fxind)
|
||||
deallocate(ansr)
|
||||
deallocate(xc)
|
||||
deallocate(ex)
|
||||
deallocate(R0)
|
||||
deallocate(R1)
|
||||
deallocate(R2)
|
||||
if (allocated(COV) ) then
|
||||
deallocate(COV)
|
||||
endif
|
||||
deallocate(h1)
|
||||
deallocate(h2)
|
||||
deallocate(hh1)
|
||||
deallocate(hh2)
|
||||
deallocate(a_up)
|
||||
deallocate(a_lo)
|
||||
stop
|
||||
!return
|
||||
|
||||
CONTAINS
|
||||
|
||||
|
||||
|
||||
SUBROUTINE INIT_LEVELS
|
||||
& (U,Ntime,N0,NIT,speed,SCIS,seed1,Nx1,Nx2,dT)
|
||||
IMPLICIT NONE
|
||||
integer, intent(out):: Ntime,N0,NIT,speed,Nx1,Nx2,SCIS,seed1
|
||||
double precision ,intent(out) :: U,dT
|
||||
|
||||
|
||||
OPEN(UNIT=14,FILE='reflev.in',STATUS= 'UNKNOWN')
|
||||
|
||||
READ (14,*) U
|
||||
READ (14,*) Ntime
|
||||
READ (14,*) N0
|
||||
READ (14,*) NIT
|
||||
READ (14,*) speed
|
||||
READ (14,*) SCIS
|
||||
READ (14,*) seed1
|
||||
|
||||
|
||||
READ (14,*) Nx1,Nx2
|
||||
READ (14,*) dT
|
||||
if (Ntime.lt.5) then
|
||||
print *,'The number of wavelength points is too small, stop'
|
||||
stop
|
||||
end if
|
||||
|
||||
CLOSE(UNIT=14)
|
||||
|
||||
RETURN
|
||||
END SUBROUTINE INIT_LEVELS
|
||||
|
||||
C******************************************************
|
||||
SUBROUTINE INIT_AMPLITUDES(h1,Nx1,h2,Nx2)
|
||||
IMPLICIT NONE
|
||||
double precision, dimension(:), intent(out) :: h1,h2
|
||||
integer, intent(in) :: Nx1,Nx2
|
||||
integer :: ix
|
||||
|
||||
|
||||
OPEN(UNIT=4,FILE='h.in',STATUS= 'UNKNOWN')
|
||||
|
||||
C
|
||||
C Reading in amplitudes
|
||||
C
|
||||
do ix=1,Nx1
|
||||
READ (4,*) H1(ix)
|
||||
enddo
|
||||
do ix=1,Nx2
|
||||
READ (4,*) H2(ix)
|
||||
enddo
|
||||
CLOSE(UNIT=4)
|
||||
|
||||
RETURN
|
||||
END SUBROUTINE INIT_AMPLITUDES
|
||||
|
||||
C**************************************************
|
||||
|
||||
C***********************************************************************
|
||||
C***********************************************************************
|
||||
|
||||
SUBROUTINE INIT_COVARIANCES(Ntime,R0,R1,R2)
|
||||
IMPLICIT NONE
|
||||
double precision, dimension(:),intent(out) :: R0,R1,R2
|
||||
integer,intent(in) :: Ntime
|
||||
integer :: i
|
||||
open (unit=1, file='Cd0.in',STATUS='unknown')
|
||||
open (unit=2, file='Cd1.in',STATUS='unknown')
|
||||
open (unit=3, file='Cd2.in',STATUS='unknown')
|
||||
|
||||
do i=1,Ntime
|
||||
read(1,*) R0(i)
|
||||
read(2,*) R1(i)
|
||||
read(3,*) R2(i)
|
||||
enddo
|
||||
close(1)
|
||||
close(2)
|
||||
close(3)
|
||||
|
||||
return
|
||||
END SUBROUTINE INIT_COVARIANCES
|
||||
|
||||
C***********************************************************************
|
||||
C***********************************************************************
|
||||
|
||||
C**********************************************************************
|
||||
|
||||
SUBROUTINE COV_INPUT(BIG,tn,ts, R0,R1,R2)
|
||||
IMPLICIT NONE
|
||||
double precision, dimension(:,:),intent(inout) :: BIG
|
||||
double precision, dimension(:),intent(in) :: R0,R1,R2
|
||||
integer ,intent(in) :: tn,ts
|
||||
integer :: i,j,Ntd1,N !=Ntdc
|
||||
double precision :: tmp
|
||||
! the order of the variables in the covariance matrix
|
||||
! are organized as follows:
|
||||
!
|
||||
! ||X(t2)..X(ts),..X(tn-1)||X'(ts) X'(t1) X'(tn)||Y1 Y2 X(ts) X(t1) X(tn)||
|
||||
! = [Xt Xd Xc]
|
||||
! where
|
||||
!
|
||||
! Xt= time points in the indicator function
|
||||
! Xd= derivatives
|
||||
! Xc=variables to condition on
|
||||
|
||||
! Computations of all covariances follows simple rules: Cov(X(t),X(s))=r(t,s),
|
||||
! then Cov(X'(t),X(s))=dr(t,s)/dt. Now for stationary X(t) we have
|
||||
! a function r(tau) such that Cov(X(t),X(s))=r(s-t) (or r(t-s) will give the same result).
|
||||
!
|
||||
! Consequently Cov(X'(t),X(s)) = -r'(s-t) = -sign(s-t)*r'(|s-t|)
|
||||
! Cov(X'(t),X'(s)) = -r''(s-t) = -r''(|s-t|)
|
||||
! Cov(X''(t),X'(s)) = r'''(s-t) = sign(s-t)*r'''(|s-t|)
|
||||
! Cov(X''(t),X(s)) = r''(s-t) = r''(|s-t|)
|
||||
! Cov(X''(t),X''(s)) = r''''(s-t) = r''''(|s-t|)
|
||||
|
||||
Ntd1=tn+1
|
||||
N=Ntd1+Nc
|
||||
do i=1,tn-2
|
||||
!cov(Xt)
|
||||
do j=i,tn-2
|
||||
BIG(i,j) = R0(j-i+1) ! cov(X(ti+1),X(tj+1))
|
||||
enddo
|
||||
!cov(Xt,Xc)
|
||||
BIG(i ,Ntd1+1) = 0.d0 !cov(X(ti+1),Y1)
|
||||
BIG(i ,Ntd1+2) = 0.d0 !cov(X(ti+1),Y2)
|
||||
BIG(i ,Ntd1+4) = R0(i+1) !cov(X(ti+1),X(t1))
|
||||
BIG(tn-1-i ,Ntd1+5) = R0(i+1) !cov(X(t.. ),X(tn))
|
||||
|
||||
!Cov(Xt,Xd)=cov(X(ti+1),x(tj)
|
||||
BIG(i,Ntd1-1) =-R1(i+1) !cov(X(ti+1),X'(t1))
|
||||
BIG(tn-1-i,Ntd1)= R1(i+1) !cov(X(ti+1),X'(tn))
|
||||
enddo
|
||||
!cov(Xd)
|
||||
BIG(Ntd1 ,Ntd1 ) = -R2(1)
|
||||
BIG(Ntd1-1,Ntd1 ) = -R2(tn) !cov(X'(t1),X'(tn))
|
||||
BIG(Ntd1-1,Ntd1-1) = -R2(1)
|
||||
BIG(Ntd1-2,Ntd1-1) = -R2(ts) !cov(X'(ts),X'(t1))
|
||||
BIG(Ntd1-2,Ntd1-2) = -R2(1)
|
||||
BIG(Ntd1-2,Ntd1 ) = -R2(tn+1-ts) !cov(X'(ts),X'(tn))
|
||||
|
||||
!cov(Xc)
|
||||
BIG(Ntd1+1,Ntd1+1) = 100.d0 ! cov(Y1 Y1)
|
||||
BIG(Ntd1+1,Ntd1+2) = 0.d0 ! cov(Y1 Y2)
|
||||
BIG(Ntd1+1,Ntd1+3) = 0.d0 ! cov(Y1 X(ts))
|
||||
BIG(Ntd1+1,Ntd1+4) = 0.d0 ! cov(Y1 X(t1))
|
||||
BIG(Ntd1+1,Ntd1+5) = 0.d0 ! cov(Y1 X(tn))
|
||||
BIG(Ntd1+2,Ntd1+2) = 100.d0 ! cov(Y2 Y2)
|
||||
BIG(Ntd1+2,Ntd1+3) = 0.d0 ! cov(Y2 X(ts))
|
||||
BIG(Ntd1+2,Ntd1+4) = 0.d0 ! cov(Y2 X(t1))
|
||||
BIG(Ntd1+2,Ntd1+5) = 0.d0 ! cov(Y2 X(tn))
|
||||
|
||||
BIG(Ntd1+3,Ntd1+3) = R0(1) ! cov(X(ts),X (ts)
|
||||
BIG(Ntd1+3,Ntd1+4) = R0(ts) ! cov(X(ts),X (t1))
|
||||
BIG(Ntd1+3,Ntd1+5) = R0(tn+1-ts) ! cov(X(ts),X (tn))
|
||||
BIG(Ntd1+4,Ntd1+4) = R0(1) ! cov(X(t1),X (t1))
|
||||
BIG(Ntd1+4,Ntd1+5) = R0(tn) ! cov(X(t1),X (tn))
|
||||
BIG(Ntd1+5,Ntd1+5) = R0(1) ! cov(X(tn),X (tn))
|
||||
|
||||
|
||||
!cov(Xd,Xc)
|
||||
BIG(Ntd1 ,Ntd1+1) = 0.d0 !cov(X'(tn),Y1)
|
||||
BIG(Ntd1 ,Ntd1+2) = 0.d0 !cov(X'(tn),Y2)
|
||||
BIG(Ntd1-1 ,Ntd1+1) = 0.d0 !cov(X'(t1),Y1)
|
||||
BIG(Ntd1-1 ,Ntd1+2) = 0.d0 !cov(X'(t1),Y2)
|
||||
BIG(Ntd1-2 ,Ntd1+1) = 0.d0 !cov(X'(ts),Y1)
|
||||
BIG(Ntd1-2 ,Ntd1+2) = 0.d0 !cov(X'(ts),Y2)
|
||||
|
||||
BIG(Ntd1 ,Ntd1+4) = R1(tn) !cov(X'(tn),X(t1))
|
||||
BIG(Ntd1 ,Ntd1+5) = 0.d0 !cov(X'(tn),X(tn))
|
||||
BIG(Ntd1-1,Ntd1+4) = 0.d0 !cov(X'(t1),X(t1))
|
||||
BIG(Ntd1-1,Ntd1+5) =-R1(tn) !cov(X'(t1),X(tn))
|
||||
BIG(Ntd1 ,Ntd1+3) = R1(tn+1-ts) !cov(X'(tn),X (ts))
|
||||
BIG(Ntd1-1,Ntd1+3) =-R1(ts) !cov(X'(t1),X (ts))
|
||||
BIG(Ntd1-2,Ntd1+3) = 0.d0 !cov(X'(ts),X (ts)
|
||||
BIG(Ntd1-2,Ntd1+4) = R1(ts) !cov(X'(ts),X (t1))
|
||||
BIG(Ntd1-2,Ntd1+5) = -R1(tn+1-ts) !cov(X'(ts),X (tn))
|
||||
|
||||
|
||||
do i=1,tn-2
|
||||
j=abs(i+1-ts)
|
||||
!cov(Xt,Xc)
|
||||
BIG(i,Ntd1+3) = R0(j+1) !cov(X(ti+1),X(ts))
|
||||
!Cov(Xt,Xd)
|
||||
if ((i+1-ts).lt.0) then
|
||||
BIG(i,Ntd1-2) = R1(j+1)
|
||||
else !cov(X(ti+1),X'(ts))
|
||||
BIG(i,Ntd1-2) = -R1(j+1)
|
||||
endif
|
||||
enddo
|
||||
|
||||
! make lower triangular part equal to upper
|
||||
do j=1,N-1
|
||||
do i=j+1,N
|
||||
tmp =BIG(j,i)
|
||||
|
||||
BIG(i,j)=tmp
|
||||
enddo
|
||||
enddo
|
||||
|
||||
C write (*,10) ((BIG(j,i),i=N+1,N+6),j=N+1,N+6)
|
||||
C 10 format(6F8.4)
|
||||
RETURN
|
||||
END SUBROUTINE COV_INPUT
|
||||
|
||||
SUBROUTINE COV_INPUT2(BIG,pt, R0,R1,R2)
|
||||
IMPLICIT NONE
|
||||
double precision, dimension(:,:), intent(out) :: BIG
|
||||
double precision, dimension(:), intent(in) :: R0,R1,R2
|
||||
integer :: pt,i,j
|
||||
! the order of the variables in the covariance matrix
|
||||
! are organized as follows;
|
||||
! X(t2)...X(tn-1) X'(t1) X'(tn) X(t1) X(tn) = [Xt Xd Xc]
|
||||
!
|
||||
! where Xd is the derivatives
|
||||
!
|
||||
! Xt= time points in the indicator function
|
||||
! Xd= derivatives
|
||||
! Xc=variables to condition on
|
||||
|
||||
!cov(Xc)
|
||||
BIG(pt+2,pt+2) = R0(1)
|
||||
BIG(pt+1,pt+1) = R0(1)
|
||||
BIG(pt+1,pt+2) = R0(pt)
|
||||
!cov(Xd)
|
||||
BIG(pt,pt) = -R2(1)
|
||||
BIG(pt-1,pt-1) = -R2(1)
|
||||
BIG(pt-1,pt) = -R2(pt)
|
||||
!cov(Xd,Xc)
|
||||
BIG(pt,pt+2) = 0.d0
|
||||
BIG(pt,pt+1) = R1(pt)
|
||||
BIG(pt-1,pt+2) = -R1(pt)
|
||||
BIG(pt-1,pt+1) = 0.d0
|
||||
|
||||
if (pt.GT.2) then
|
||||
!cov(Xt)
|
||||
do i=1,pt-2
|
||||
do j=i,pt-2
|
||||
BIG(i,j) = R0(j-i+1)
|
||||
enddo
|
||||
enddo
|
||||
!cov(Xt,Xc)
|
||||
do i=1,pt-2
|
||||
BIG(i,pt+1) = R0(i+1)
|
||||
BIG(pt-1-i,pt+2) = R0(i+1)
|
||||
enddo
|
||||
!Cov(Xt,Xd)=cov(X(ti+1),x(tj))
|
||||
do i=1,pt-2
|
||||
BIG(i,pt-1) = -R1(i+1)
|
||||
BIG(pt-1-i,pt)= R1(i+1)
|
||||
enddo
|
||||
endif
|
||||
|
||||
|
||||
! make lower triangular part equal to upper
|
||||
do j=1,pt+1
|
||||
do i=j+1,pt+2
|
||||
BIG(i,j)=BIG(j,i)
|
||||
enddo
|
||||
enddo
|
||||
C write (*,10) ((BIG(j,i),i=N+1,N+6),j=N+1,N+6)
|
||||
C 10 format(6F8.4)
|
||||
RETURN
|
||||
END SUBROUTINE COV_INPUT2
|
||||
|
||||
|
||||
END PROGRAM sp2tccpdf
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
@ -1,440 +0,0 @@
|
||||
PROGRAM sp2tcpdf
|
||||
C***********************************************************************
|
||||
C This program computes: *
|
||||
C *
|
||||
C density of T_i, for Ac <=h, in a gaussian process i.e. *
|
||||
C *
|
||||
C half wavelength (up-crossing to downcrossing) for crests <h *
|
||||
C or half wavelength (down-crossing to upcrossing) for trough >h *
|
||||
C***********************************************************************
|
||||
use GLOBALDATA, only : Nt,Nj,Nd,Nc,Ntd,Ntdc,NI,Mb,
|
||||
&NIT,Nx,TWOPI,XSPLT,SCIS,NSIMmax,COV
|
||||
use rind
|
||||
IMPLICIT NONE
|
||||
double precision, dimension(:,:),allocatable :: BIG
|
||||
double precision, dimension(:,:),allocatable :: ansr
|
||||
double precision, dimension(: ),allocatable :: ex,CY
|
||||
double precision, dimension(:,:),allocatable :: xc
|
||||
double precision, dimension(: ),allocatable :: fxind,h
|
||||
double precision, dimension(: ),allocatable :: R0,R1,R2,R3,R4
|
||||
double precision ::CC,U,XddInf,XdInf,XtInf
|
||||
double precision, dimension(2,6) :: a_up=0.d0,a_lo=0.d0
|
||||
integer , dimension(: ),allocatable :: seed
|
||||
integer ,dimension(7) :: indI
|
||||
integer :: Nstart,Ntime,tn,ts,speed,ph,def,seed1,seed_size,icy
|
||||
integer ::it1,it2
|
||||
double precision :: ds,dT ! lag spacing for covariances
|
||||
! DIGITAL:
|
||||
! f90 -o ~/WAT/V1/sp2tcpdf.exe rind44.f sp2tcpdf.f
|
||||
|
||||
!print *,'enter sp2thpdf'
|
||||
CALL INIT_LEVELS(U,def,Ntime,Nstart,NIT,speed,Nx,dT)
|
||||
!print *,'U,def,Ntime,Nstart,NIT,speed,SCIS,seed1,Nx,dT'
|
||||
!print *,U,def,Ntime,Nstart,NIT,speed,SCIS,seed1,Nx,dT
|
||||
!XSPLT=1.5d0
|
||||
if (SCIS.GT.0) then
|
||||
allocate(COV(1:Nx))
|
||||
call random_seed(SIZE=seed_size)
|
||||
allocate(seed(seed_size))
|
||||
call random_seed(GET=seed(1:seed_size)) ! get current seed
|
||||
seed(1)=seed1 ! change seed
|
||||
call random_seed(PUT=seed(1:seed_size))
|
||||
deallocate(seed)
|
||||
endif
|
||||
CALL INITDATA(speed)
|
||||
!print *,ntime,speed,u,NIT
|
||||
allocate(R0(1:Ntime+1))
|
||||
allocate(R1(1:Ntime+1))
|
||||
allocate(R2(1:Ntime+1))
|
||||
if (abs(def).GT.1) THEN
|
||||
!allocate(h(1:Nx))
|
||||
allocate(R3(1:Ntime+1))
|
||||
allocate(R4(1:Ntime+1))
|
||||
!CALL INIT_AMPLITUDES(h,def,Nx)
|
||||
endif
|
||||
allocate(h(1:Nx))
|
||||
CALL INIT_AMPLITUDES(h,def,Nx)
|
||||
CALL INIT_COVARIANCES(Ntime,def,R0,R1,R2,R3,R4)
|
||||
|
||||
print *,'Nx',Nx
|
||||
|
||||
Nj=0
|
||||
indI(1)=0
|
||||
C ***** The bound 'infinity' is set to 10*sigma *****
|
||||
XdInf=10.d0*SQRT(-R2(1))
|
||||
XtInf=10.d0*SQRT(R0(1))
|
||||
print *,'XdInf,XtInf'
|
||||
print *,XdInf,XtInf
|
||||
! normalizing constant
|
||||
CC=TWOPI*SQRT(-R0(1)/R2(1))*exp(u*u/(2.d0*R0(1)) )
|
||||
!fy(h)
|
||||
allocate(CY(1:Nx))
|
||||
do icy=1,Nx
|
||||
CY(icy)=exp(-0.5*h(icy)*h(icy)/100)/(10*sqrt(twopi))
|
||||
enddo
|
||||
|
||||
allocate(ansr(1:Ntime,1:Nx))
|
||||
ansr=0.d0
|
||||
allocate(fxind(1:Nx))
|
||||
fxind=0.d0 !this is not needed
|
||||
|
||||
NI=4; Nd=2
|
||||
Nc=3; Mb=2
|
||||
allocate(BIG(1:Ntime+Nc,1:Ntime+Nc))
|
||||
allocate(xc(1:Nc,1:Nx))
|
||||
allocate(ex(1:Ntime+Nc))
|
||||
ex=0.d0
|
||||
!print *,'nc',Nc,Nx
|
||||
xc(1,1:Nx)=h(1:Nx)
|
||||
print *,'xc',h(1)
|
||||
print *,'test',def;
|
||||
xc(2,1:Nx)=u
|
||||
xc(3,1:Nx)=u
|
||||
if (def.GT.0) then
|
||||
a_up(1,1)=u !+XtInf
|
||||
a_lo(1,1)=u
|
||||
a_up(1,2)=XdInf
|
||||
a_lo(1,3)=-XdInf
|
||||
a_up(2,1)=1.d0
|
||||
else
|
||||
a_up(1,1)=u
|
||||
a_lo(1,1)=u !-XtInf
|
||||
a_lo(1,2)=-XdInf
|
||||
a_up(1,3)= XdInf
|
||||
a_lo(2,1)=1.d0
|
||||
print *,'a_lo',a_lo(2,1)
|
||||
endif
|
||||
!print *,'Nstart',Nstart
|
||||
Nstart=MAX(3,Nstart)
|
||||
!print *,'Nstart',Nstart
|
||||
if (SCIS.GT.0) then
|
||||
open (unit=11, file='COV.out', STATUS='unknown')
|
||||
write(11,*) 0.d0
|
||||
endif
|
||||
do Ntd=Nstart,Ntime
|
||||
!CALL COV_INPUT2(BIG,Ntd, R0,R1,R2)
|
||||
Ntdc=Ntd+Nc;
|
||||
CALL COV_INPUT(BIG,Ntd,-1,R0,R1,R2,R3,R4) ! positive wave period
|
||||
Nt=Ntd-Nd;
|
||||
indI(2)=Nt;
|
||||
indI(3)=Nt+1;
|
||||
indI(4)=Ntd;
|
||||
!Ntdc=Ntd+Nc;
|
||||
|
||||
CALL RINDD(fxind,Big(1:Ntdc,1:Ntdc),ex(1:Ntdc),
|
||||
& xc,indI,a_lo,a_up)
|
||||
! print *,'test',fxind/CY(1:Nx)
|
||||
do icy=1,Nx
|
||||
ansr(Ntd,icy)=fxind(icy)*CC/CY(icy)
|
||||
enddo
|
||||
if (SCIS.GT.0) then
|
||||
write(11,*) COV(1) ! save coefficient of variation
|
||||
endif
|
||||
print *,'Ready: ',Ntd,' of ',Ntime
|
||||
enddo
|
||||
goto 300
|
||||
300 open (unit=11, file='dens.out', STATUS='unknown')
|
||||
|
||||
!print *, ansr
|
||||
do ts=1,Ntime
|
||||
do ph=1,Nx
|
||||
write(11,*) ansr(ts,ph)
|
||||
! write(11,111) ansr(ts,ph)
|
||||
enddo
|
||||
enddo
|
||||
!111 FORMAT(2x,F12.8)
|
||||
close(11)
|
||||
900 deallocate(big)
|
||||
deallocate(fxind)
|
||||
deallocate(ansr)
|
||||
deallocate(xc)
|
||||
deallocate(ex)
|
||||
deallocate(R0)
|
||||
deallocate(R1)
|
||||
deallocate(R2)
|
||||
if (allocated(COV) ) then
|
||||
deallocate(COV)
|
||||
endif
|
||||
|
||||
if (allocated(R3)) then
|
||||
deallocate(R3)
|
||||
deallocate(R4)
|
||||
deallocate(h)
|
||||
ENDIF
|
||||
stop
|
||||
!return
|
||||
|
||||
CONTAINS
|
||||
|
||||
|
||||
|
||||
SUBROUTINE INIT_LEVELS
|
||||
& (U,def,Ntime,Nstart,NIT,speed,Nx,dT)
|
||||
IMPLICIT NONE
|
||||
integer, intent(out):: def,Ntime,Nstart,NIT,speed,Nx
|
||||
double precision ,intent(out) :: U,dT
|
||||
|
||||
|
||||
OPEN(UNIT=14,FILE='reflev.in',STATUS= 'UNKNOWN')
|
||||
READ (14,*) U
|
||||
READ (14,*) def
|
||||
READ (14,*) Ntime
|
||||
READ (14,*) Nstart
|
||||
READ (14,*) NIT
|
||||
READ (14,*) speed
|
||||
READ (14,*) SCIS
|
||||
READ (14,*) seed1
|
||||
READ (14,*) Nx
|
||||
print *,'def',def
|
||||
if (abs(def).GT.1) then
|
||||
READ (14,*) dT
|
||||
if (Ntime.lt.3) then
|
||||
print *,'The number of wavelength points is too small, stop'
|
||||
stop
|
||||
end if
|
||||
else
|
||||
if (Ntime.lt.2) then
|
||||
print *,'The number of wavelength points is too small, stop'
|
||||
stop
|
||||
end if
|
||||
endif
|
||||
CLOSE(UNIT=14)
|
||||
|
||||
RETURN
|
||||
END SUBROUTINE INIT_LEVELS
|
||||
|
||||
C******************************************************
|
||||
SUBROUTINE INIT_AMPLITUDES(h,def,Nx)
|
||||
IMPLICIT NONE
|
||||
double precision, dimension(:), intent(out) :: h
|
||||
integer, intent(in) :: def
|
||||
integer, intent(in) :: Nx
|
||||
integer :: ix
|
||||
|
||||
|
||||
OPEN(UNIT=4,FILE='h.in',STATUS= 'UNKNOWN')
|
||||
|
||||
C
|
||||
C Reading in amplitudes
|
||||
C
|
||||
do ix=1,Nx
|
||||
READ (4,*) H(ix)
|
||||
enddo
|
||||
CLOSE(UNIT=4)
|
||||
!if (def.LT.0) THEN
|
||||
! H=-H
|
||||
!endif
|
||||
|
||||
RETURN
|
||||
END SUBROUTINE INIT_AMPLITUDES
|
||||
|
||||
C**************************************************
|
||||
|
||||
C***********************************************************************
|
||||
C***********************************************************************
|
||||
|
||||
SUBROUTINE INIT_COVARIANCES(Ntime,def,R0,R1,R2,R3,R4)
|
||||
IMPLICIT NONE
|
||||
double precision, dimension(:),intent(out) :: R0,R1,R2
|
||||
double precision, dimension(:),intent(out) :: R3,R4
|
||||
integer,intent(in) :: Ntime,def
|
||||
integer :: i
|
||||
open (unit=1, file='Cd0.in',STATUS='unknown')
|
||||
open (unit=2, file='Cd1.in',STATUS='unknown')
|
||||
open (unit=3, file='Cd2.in',STATUS='unknown')
|
||||
|
||||
do i=1,Ntime
|
||||
read(1,*) R0(i)
|
||||
read(2,*) R1(i)
|
||||
read(3,*) R2(i)
|
||||
enddo
|
||||
close(1)
|
||||
close(2)
|
||||
close(3)
|
||||
|
||||
if (abs(def).GT.1) then
|
||||
open (unit=4, file='Cd3.in',STATUS='unknown')
|
||||
open (unit=5, file='Cd4.in',STATUS='unknown')
|
||||
|
||||
do i=1,Ntime
|
||||
read(4,*) R3(i)
|
||||
read(5,*) R4(i)
|
||||
enddo
|
||||
|
||||
close(4)
|
||||
close(5)
|
||||
endif
|
||||
return
|
||||
END SUBROUTINE INIT_COVARIANCES
|
||||
|
||||
C***********************************************************************
|
||||
C***********************************************************************
|
||||
|
||||
C**********************************************************************
|
||||
|
||||
SUBROUTINE COV_INPUT(BIG,tn,ts, R0,R1,R2,R3,R4)
|
||||
IMPLICIT NONE
|
||||
double precision, dimension(:,:),intent(inout) :: BIG
|
||||
double precision, dimension(:),intent(in) :: R0,R1,R2
|
||||
double precision, dimension(:),intent(in) :: R3,R4
|
||||
integer ,intent(in) :: tn,ts
|
||||
integer :: i,j,shft,Ntd1,N !=Ntdc
|
||||
double precision :: tmp
|
||||
! the order of the variables in the covariance matrix
|
||||
! are organized as follows:
|
||||
! For ts>1:
|
||||
! X(t2)..X(ts),..X(tn-1) X''(ts) X'(t1) X'(tn) X(ts) X(t1) X(tn) X'(ts)
|
||||
! = [Xt Xd Xc]
|
||||
!
|
||||
! For ts<=1:
|
||||
! X(t2)..,..X(tn-1) X'(t1) X'(tn) Y X(t1) X(tn)
|
||||
! = [Xt Xd Xc]
|
||||
!Add Y Condition : Y=h
|
||||
|
||||
! where
|
||||
!
|
||||
! Xt= time points in the indicator function
|
||||
! Xd= derivatives
|
||||
! Xc=variables to condition on
|
||||
|
||||
if (ts.LE.1) THEN
|
||||
Ntd1=tn
|
||||
N=Ntd1+Nc;
|
||||
shft=0 ! def=1 want only crest period Tc
|
||||
else
|
||||
Ntd1=tn+1
|
||||
N=Ntd1+4
|
||||
shft=1 ! def=2 or 3 want Tc Ac or Tcf, Ac
|
||||
endif
|
||||
|
||||
do i=1,tn-2
|
||||
!cov(Xt)
|
||||
do j=i,tn-2
|
||||
BIG(i,j) = R0(j-i+1) ! cov(X(ti+1),X(tj+1))
|
||||
enddo
|
||||
!cov(Xt,Xc)
|
||||
BIG(i ,Ntd1+1+shft) = 0.d0 !cov(X(ti+1),Y)
|
||||
BIG(i ,Ntd1+2+shft) = R0(i+1) !cov(X(ti+1),X(t1))
|
||||
BIG(tn-1-i ,Ntd1+3+shft) = R0(i+1) !cov(X(t.. ),X(tn))
|
||||
!Cov(Xt,Xd)=cov(X(ti+1),x(tj)
|
||||
BIG(i,Ntd1-1) =-R1(i+1) !cov(X(ti+1),X' (t1))
|
||||
BIG(tn-1-i,Ntd1)= R1(i+1) !cov(X(ti+1),X' (tn))
|
||||
enddo
|
||||
!call echo(big(1:tn,1:tn),tn)
|
||||
!cov(Xd)
|
||||
BIG(Ntd1 ,Ntd1 ) = -R2(1)
|
||||
BIG(Ntd1-1,Ntd1 ) = -R2(tn) !cov(X'(t1),X'(tn))
|
||||
BIG(Ntd1-1,Ntd1-1) = -R2(1)
|
||||
|
||||
!cov(Xc)
|
||||
!print *,'t'
|
||||
BIG(Ntd1+1+shft,Ntd1+1+shft) = 100.d0!100.d0 ! cov(Y,Y)
|
||||
BIG(Ntd1+1+shft,Ntd1+2+shft) = 0.d0
|
||||
BIG(Ntd1+1+shft,Ntd1+3+shft) = 0.d0
|
||||
BIG(Ntd1+2+shft,Ntd1+2+shft) = R0(1) ! cov(X(t1),X (t1))
|
||||
BIG(Ntd1+2+shft,Ntd1+3+shft) = R0(tn) ! cov(X(t1),X (tn))
|
||||
BIG(Ntd1+3+shft,Ntd1+3+shft) = R0(1) ! cov(X(tn),X (tn))
|
||||
!cov(Xd,Xc)
|
||||
BIG(Ntd1 ,Ntd1+1+shft) = 0.d0 !cov(X'(tn),Y)
|
||||
BIG(Ntd1 ,Ntd1+2+shft) = R1(tn) !cov(X'(tn),X(t1))
|
||||
BIG(Ntd1 ,Ntd1+3+shft) = 0.d0 !cov(X'(tn),X(tn))
|
||||
BIG(Ntd1-1,Ntd1+1+shft) = 0.d0 !cov(X'(t1),Y)
|
||||
BIG(Ntd1-1,Ntd1+2+shft) = 0.d0 !cov(X'(t1),X(t1))
|
||||
BIG(Ntd1-1,Ntd1+3+shft) =-R1(tn) !cov(X'(t1),X(tn))
|
||||
|
||||
|
||||
!call echo(big(1:N,1:N),N)
|
||||
! make lower triangular part equal to upper
|
||||
do j=1,N-1
|
||||
do i=j+1,N
|
||||
tmp =BIG(j,i)
|
||||
BIG(i,j)=tmp
|
||||
enddo
|
||||
!call echo(big(1:N,1:N),N)
|
||||
enddo
|
||||
!if (tn.eq.3) then
|
||||
!do j=1,N
|
||||
! do i=j,N
|
||||
! print *,'test',j,i,BIG(j,i)
|
||||
! enddo
|
||||
!call echo(big(1:N,1:N),N)
|
||||
!enddo
|
||||
!endif
|
||||
!call echo(big(1:N,1:N),N)
|
||||
|
||||
C write (*,10) ((BIG(j,i),i=N+1,N+6),j=N+1,N+6)
|
||||
C 10 format(6F8.4)
|
||||
RETURN
|
||||
END SUBROUTINE COV_INPUT
|
||||
|
||||
SUBROUTINE COV_INPUT2(BIG,pt, R0,R1,R2)
|
||||
IMPLICIT NONE
|
||||
double precision, dimension(:,:), intent(out) :: BIG
|
||||
double precision, dimension(:), intent(in) :: R0,R1,R2
|
||||
integer :: pt,i,j
|
||||
! the order of the variables in the covariance matrix
|
||||
! are organized as follows;
|
||||
! X(t2)...X(tn-1) X'(t1) X'(tn) X(t1) X(tn) = [Xt Xd Xc]
|
||||
!
|
||||
! where Xd is the derivatives
|
||||
!
|
||||
! Xt= time points in the indicator function
|
||||
! Xd= derivatives
|
||||
! Xc=variables to condition on
|
||||
|
||||
!cov(Xc)
|
||||
BIG(pt+2,pt+2) = R0(1)
|
||||
BIG(pt+1,pt+1) = R0(1)
|
||||
BIG(pt+1,pt+2) = R0(pt)
|
||||
!cov(Xd)
|
||||
BIG(pt,pt) = -R2(1)
|
||||
BIG(pt-1,pt-1) = -R2(1)
|
||||
BIG(pt-1,pt) = -R2(pt)
|
||||
!cov(Xd,Xc)
|
||||
BIG(pt,pt+2) = 0.d0
|
||||
BIG(pt,pt+1) = R1(pt)
|
||||
BIG(pt-1,pt+2) = -R1(pt)
|
||||
BIG(pt-1,pt+1) = 0.d0
|
||||
|
||||
if (pt.GT.2) then
|
||||
!cov(Xt)
|
||||
do i=1,pt-2
|
||||
do j=i,pt-2
|
||||
BIG(i,j) = R0(j-i+1)
|
||||
enddo
|
||||
enddo
|
||||
!cov(Xt,Xc)
|
||||
do i=1,pt-2
|
||||
BIG(i,pt+1) = R0(i+1)
|
||||
BIG(pt-1-i,pt+2) = R0(i+1)
|
||||
enddo
|
||||
!Cov(Xt,Xd)=cov(X(ti+1),x(tj))
|
||||
do i=1,pt-2
|
||||
BIG(i,pt-1) = -R1(i+1)
|
||||
BIG(pt-1-i,pt)= R1(i+1)
|
||||
enddo
|
||||
endif
|
||||
|
||||
|
||||
! make lower triangular part equal to upper
|
||||
do j=1,pt+1
|
||||
do i=j+1,pt+2
|
||||
BIG(i,j)=BIG(j,i)
|
||||
enddo
|
||||
enddo
|
||||
C write (*,10) ((BIG(j,i),i=N+1,N+6),j=N+1,N+6)
|
||||
C 10 format(6F8.4)
|
||||
RETURN
|
||||
END SUBROUTINE COV_INPUT2
|
||||
|
||||
|
||||
END PROGRAM sp2tcpdf
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
@ -1,569 +0,0 @@
|
||||
PROGRAM sp2thpdf
|
||||
!***********************************************************************
|
||||
! This program computes: *
|
||||
! *
|
||||
! density of S_i,Hi,T_i in a gaussian process i.e. *
|
||||
! *
|
||||
! quart wavelength (up-crossing to crest) and crest amplitude *
|
||||
!
|
||||
! def = 1, gives half wave period, Tc (default).
|
||||
! -1, gives half wave period, Tt.
|
||||
! 2, gives half wave period and wave crest amplitude (Tc,Ac).
|
||||
! -2, gives half wave period and wave trough amplitude (Tt,At).
|
||||
! 3, gives crest front period and wave crest amplitude (Tcf,Ac).
|
||||
! -3, gives trough back period and wave trough amplitude (Ttb,At).
|
||||
! 4, gives minimum of crest front/back period and wave crest
|
||||
! amplitude (max(Tcf,Tcb),Ac).
|
||||
! -4, gives minimum of trough front/back period and wave trough
|
||||
! amplitude (max(Ttf,Ttb),At).
|
||||
!***********************************************************************
|
||||
use GLOBALDATA, only : Nt,Nj,Nd,Nc,Ntd,Ntdc,NI,Mb,
|
||||
& NIT,Nx,TWOPI,XSPLT,SCIS,NSIMmax,COV
|
||||
use rind
|
||||
IMPLICIT NONE
|
||||
double precision, dimension(:,:),allocatable :: BIG
|
||||
double precision, dimension(:,:),allocatable :: ansr
|
||||
double precision, dimension(: ),allocatable :: ex
|
||||
double precision, dimension(:,:),allocatable :: xc
|
||||
double precision, dimension(: ),allocatable :: fxind,h
|
||||
double precision, dimension(: ),allocatable :: R0,R1,R2,R3,R4
|
||||
double precision ::CC,U,XddInf,XdInf,XtInf
|
||||
double precision, dimension(2,6) :: a_up=0.d0,a_lo=0.d0
|
||||
integer , dimension(: ),allocatable :: seed
|
||||
integer ,dimension(7) :: indI
|
||||
integer :: Nstart,Ntime,tn,ts,speed,ph,def,seed1,seed_size
|
||||
double precision :: ds,dT ! lag spacing for covariances
|
||||
! DIGITAL:
|
||||
! f90 -g2 -C -automatic -o ../wave/alpha/sp2thpdf.exe rind44.f sp2thpdf.f
|
||||
! SOLARIS:
|
||||
!f90 -g -O -w3 -Bdynamic -fixed -o ../wave/sol2/sp2thpdf.exe rind44.f sp2thpdf.f
|
||||
! linux:
|
||||
! f90 -gline -Nl126 -C -o sp2thpdf.exe rind45.f sp2thpdf.f
|
||||
! HP700
|
||||
!f90 -g -C -o ../exec/hp700/sp2thpdf.exe rind45.f sp2thpdf.f
|
||||
!f90 -g -C +check=all +FPVZID -o ../exec/hp700/sp2thpdf2.exe rind45.f sp2thpdf.f
|
||||
|
||||
|
||||
!print *,'enter sp2thpdf'
|
||||
CALL INIT_LEVELS(U,def,Ntime,Nstart,NIT,speed,SCIS,seed1,Nx,dT)
|
||||
!print *,'U,def,Ntime,Nstart,NIT,speed,SCIS,seed1,Nx,dT'
|
||||
!print *,U,def,Ntime,Nstart,NIT,speed,SCIS,seed1,Nx,dT
|
||||
|
||||
!XSPLT=1.5d0
|
||||
if (SCIS.GT.0) then
|
||||
allocate(COV(1:Nx))
|
||||
call random_seed(SIZE=seed_size)
|
||||
allocate(seed(seed_size))
|
||||
call random_seed(GET=seed(1:seed_size)) ! get current seed
|
||||
seed(1)=seed1 ! change seed
|
||||
call random_seed(PUT=seed(1:seed_size))
|
||||
deallocate(seed)
|
||||
endif
|
||||
CALL INITDATA(speed)
|
||||
!print *,ntime,speed,u,NIT
|
||||
allocate(R0(1:Ntime+1))
|
||||
allocate(R1(1:Ntime+1))
|
||||
allocate(R2(1:Ntime+1))
|
||||
if (abs(def).GT.1) THEN
|
||||
allocate(h(1:Nx))
|
||||
allocate(R3(1:Ntime+1))
|
||||
allocate(R4(1:Ntime+1))
|
||||
|
||||
CALL INIT_AMPLITUDES(h,def,Nx)
|
||||
endif
|
||||
CALL INIT_COVARIANCES(Ntime,def,R0,R1,R2,R3,R4)
|
||||
|
||||
!print *,'Nx',Nx
|
||||
|
||||
Nj=0
|
||||
indI(1)=0
|
||||
C ***** The bound 'infinity' is set to 10*sigma *****
|
||||
XdInf=10.d0*SQRT(-R2(1))
|
||||
XtInf=10.d0*SQRT(R0(1))
|
||||
!print *,'XdInf,XtInf'
|
||||
!print *,XdInf,XtInf
|
||||
! normalizing constant
|
||||
CC=TWOPI*SQRT(-R0(1)/R2(1))*exp(u*u/(2.d0*R0(1)) )
|
||||
if (abs(def).EQ.4) CC=2.d0*CC
|
||||
allocate(ansr(1:Ntime,1:Nx))
|
||||
ansr=0.d0
|
||||
allocate(fxind(1:Nx))
|
||||
!fxind=0.d0 this is not needed
|
||||
|
||||
if (abs(def).GT.1) then
|
||||
GOTO 200
|
||||
endif
|
||||
NI=4; Nd=2
|
||||
Nc=2; Mb=1
|
||||
Nx=1
|
||||
allocate(BIG(1:Ntime+Nc,1:Ntime+Nc))
|
||||
allocate(xc(1:Nc,1:Nx))
|
||||
allocate(ex(1:Ntime+Nc))
|
||||
ex=0.d0
|
||||
xc(1,1)=u
|
||||
xc(2,1)=u
|
||||
|
||||
if (def.GT.0) then
|
||||
a_up(1,1)=u+XtInf
|
||||
a_lo(1,1)=u
|
||||
a_up(1,2)=XdInf
|
||||
a_lo(1,3)=-XdInf
|
||||
else
|
||||
a_up(1,1)=u
|
||||
a_lo(1,1)=u-XtInf
|
||||
a_lo(1,2)=-XdInf
|
||||
a_up(1,3)= XdInf
|
||||
endif
|
||||
!print *,'Nstart',Nstart
|
||||
Nstart=MAX(2,Nstart)
|
||||
!print *,'Nstart',Nstart
|
||||
if (SCIS.GT.0) then
|
||||
open (unit=11, file='COV.out', STATUS='unknown')
|
||||
write(11,*) 0.d0
|
||||
endif
|
||||
do Ntd=Nstart,Ntime
|
||||
!CALL COV_INPUT2(BIG,Ntd, R0,R1,R2)
|
||||
CALL COV_INPUT(BIG,Ntd,-1,R0,R1,R2,R3,R4) ! positive wave period
|
||||
Nt=Ntd-Nd;
|
||||
indI(2)=Nt;
|
||||
indI(3)=Nt+1;
|
||||
indI(4)=Ntd;
|
||||
Ntdc=Ntd+Nc;
|
||||
!if (SCIS.gt.0) then
|
||||
! if (SCIS.EQ.2) then
|
||||
! Nj=max(Nt,0)
|
||||
! else
|
||||
! Nj=min(max(Nt-5, 0),0)
|
||||
! endif
|
||||
!endif
|
||||
!Ex=0.d0
|
||||
!CALL echo(BIG(1:Ntdc,1:min(7,Ntdc)),Ntdc)
|
||||
CALL RINDD(fxind,Big(1:Ntdc,1:Ntdc),ex(1:Ntdc),
|
||||
& xc,indI,a_lo,a_up)
|
||||
ansr(Ntd,1)=fxind(1)*CC
|
||||
if (SCIS.GT.0) then
|
||||
write(11,*) COV(1) ! save coefficient of variation
|
||||
endif
|
||||
print *,'Ready: ',Ntd,' of ',Ntime
|
||||
enddo
|
||||
if (SCIS.GT.0) then
|
||||
close(11)
|
||||
endif
|
||||
goto 300
|
||||
200 continue
|
||||
XddInf=10.d0*SQRT(R4(1))
|
||||
NI=7; Nd=3
|
||||
Nc=4; Mb=2
|
||||
allocate(BIG(1:Ntime+Nc+1,1:Ntime+Nc+1))
|
||||
ALLOCATE(xc(1:Nc,1:Nx))
|
||||
allocate(ex(1:Ntime+Nc+1))
|
||||
|
||||
ex=0.d0
|
||||
xc(1,1:Nx)=h
|
||||
xc(2,1:Nx)=u
|
||||
xc(3,1:Nx)=u
|
||||
xc(4,1:Nx)=0.d0
|
||||
|
||||
if (def.GT.0) then
|
||||
a_up(2,1)=1.d0 !*h
|
||||
a_lo(1,1)=u
|
||||
a_up(1,2)=XtInf ! X(ts) is redundant
|
||||
a_lo(1,2)=-Xtinf
|
||||
a_up(2,2)=1.d0 ! *h
|
||||
a_lo(2,2)=1.d0 ! *h
|
||||
a_up(2,3)=1.d0 !*h
|
||||
a_lo(1,3)=u
|
||||
|
||||
a_lo(1,4)=-XddInf
|
||||
a_up(1,5)= XdInf
|
||||
a_lo(1,6)=-XdInf
|
||||
else !def<0
|
||||
a_up(1,1)=u
|
||||
a_lo(2,1)=1.d0 !*h
|
||||
a_up(1,2)=XtInf ! X(ts) is redundant
|
||||
a_lo(1,2)=-Xtinf
|
||||
a_up(2,2)=1.d0 ! *h
|
||||
a_lo(2,2)=1.d0 ! *h
|
||||
a_up(1,3)=u
|
||||
a_lo(2,3)=1.d0 !*h
|
||||
|
||||
a_up(1,4)=XddInf
|
||||
a_lo(1,5)=-XdInf
|
||||
a_up(1,6)=XdInf
|
||||
endif
|
||||
|
||||
Nstart=MAX(Nstart,3)
|
||||
do tn=Nstart,Ntime,1
|
||||
Ntd=tn+1
|
||||
Nt=Ntd-Nd
|
||||
Ntdc=Ntd+Nc
|
||||
indI(4)=Nt
|
||||
indI(5)=Nt+1
|
||||
indI(6)=Nt+2
|
||||
indI(7)=Ntd
|
||||
if (SCIS.gt.0) then
|
||||
if (SCIS.EQ.2) then
|
||||
Nj=max(Nt,0)
|
||||
else
|
||||
Nj=min(max(Nt-5, 0),0)
|
||||
endif
|
||||
endif
|
||||
do ts=2,FLOOR(DBLE(tn+1)/2.d0)
|
||||
!print *,'ts,tn' ,ts,tn
|
||||
CALL COV_INPUT(Big(1:Ntdc,1:Ntdc),tn,ts,R0,R1,R2,R3,R4) ! positive wave period
|
||||
indI(2)=ts-2
|
||||
indI(3)=ts-1
|
||||
!CALL echo(BIG(1:Ntdc,1:min(7,Ntdc)),Ntdc)
|
||||
!print *,'sp call rind'
|
||||
CALL RINDD(fxind,Big(1:Ntdc,1:Ntdc),ex(1:Ntdc),
|
||||
& xc,indI,a_lo,a_up)
|
||||
!CALL echo(BIG(1:Ntdc,1:min(7,Ntdc)),Ntdc)
|
||||
!print *,'sp rind finished',fxind
|
||||
!goto 900
|
||||
if (abs(def).LT.3) THEN
|
||||
if (ts .EQ.tn-ts+1) then
|
||||
ds=dt
|
||||
else
|
||||
ds=2.d0*dt
|
||||
endif
|
||||
ansr(tn,1:Nx)=ansr(tn,1:Nx)+fxind*CC*ds
|
||||
else
|
||||
ansr(ts,1:Nx)=ansr(ts,1:Nx)+fxind*CC*dT
|
||||
if ((ts.LT.tn-ts+1).and. (abs(def).lt.4)) THEN
|
||||
ansr(tn-ts+1,1:Nx)=ansr(tn-ts+1,1:Nx)+fxind*CC*dT ! exploiting the symmetry
|
||||
endif
|
||||
endif
|
||||
enddo ! ts
|
||||
print *,'Ready: ',tn,' of ',Ntime
|
||||
|
||||
enddo !tn
|
||||
!print *,'ansr',ansr
|
||||
300 open (unit=11, file='dens.out', STATUS='unknown')
|
||||
!print *, ansr
|
||||
do ts=1,Ntime
|
||||
do ph=1,Nx
|
||||
write(11,*) ansr(ts,ph)
|
||||
! write(11,111) ansr(ts,ph)
|
||||
enddo
|
||||
enddo
|
||||
!111 FORMAT(2x,F12.8)
|
||||
close(11)
|
||||
900 deallocate(big)
|
||||
deallocate(fxind)
|
||||
deallocate(ansr)
|
||||
deallocate(xc)
|
||||
deallocate(ex)
|
||||
deallocate(R0)
|
||||
deallocate(R1)
|
||||
deallocate(R2)
|
||||
if (allocated(COV) ) then
|
||||
deallocate(COV)
|
||||
endif
|
||||
if (allocated(R3)) then
|
||||
deallocate(R3)
|
||||
deallocate(R4)
|
||||
deallocate(h)
|
||||
ENDIF
|
||||
stop
|
||||
!return
|
||||
|
||||
CONTAINS
|
||||
|
||||
|
||||
|
||||
SUBROUTINE INIT_LEVELS
|
||||
& (U,def,Ntime,Nstart,NIT,speed,SCIS,seed1,Nx,dT)
|
||||
IMPLICIT NONE
|
||||
integer, intent(out):: def,Ntime,Nstart,NIT,speed,Nx,SCIS,seed1
|
||||
double precision ,intent(out) :: U,dT
|
||||
|
||||
|
||||
OPEN(UNIT=14,FILE='reflev.in',STATUS= 'UNKNOWN')
|
||||
READ (14,*) U
|
||||
READ (14,*) def
|
||||
READ (14,*) Ntime
|
||||
READ (14,*) Nstart
|
||||
READ (14,*) NIT
|
||||
READ (14,*) speed
|
||||
READ (14,*) SCIS
|
||||
READ (14,*) seed1
|
||||
|
||||
if (abs(def).GT.1) then
|
||||
READ (14,*) Nx
|
||||
READ (14,*) dT
|
||||
if (Ntime.lt.3) then
|
||||
print *,'The number of wavelength points is too small, stop'
|
||||
stop
|
||||
end if
|
||||
else
|
||||
Nx=1
|
||||
if (Ntime.lt.2) then
|
||||
print *,'The number of wavelength points is too small, stop'
|
||||
stop
|
||||
end if
|
||||
endif
|
||||
CLOSE(UNIT=14)
|
||||
|
||||
RETURN
|
||||
END SUBROUTINE INIT_LEVELS
|
||||
|
||||
C******************************************************
|
||||
SUBROUTINE INIT_AMPLITUDES(h,def,Nx)
|
||||
IMPLICIT NONE
|
||||
double precision, dimension(:), intent(out) :: h
|
||||
integer, intent(in) :: def
|
||||
integer, intent(in) :: Nx
|
||||
integer :: ix
|
||||
|
||||
|
||||
OPEN(UNIT=4,FILE='h.in',STATUS= 'UNKNOWN')
|
||||
|
||||
C
|
||||
C Reading in amplitudes
|
||||
C
|
||||
do ix=1,Nx
|
||||
READ (4,*) H(ix)
|
||||
enddo
|
||||
CLOSE(UNIT=4)
|
||||
!if (def.LT.0) THEN
|
||||
! H=-H
|
||||
!endif
|
||||
|
||||
RETURN
|
||||
END SUBROUTINE INIT_AMPLITUDES
|
||||
|
||||
C**************************************************
|
||||
|
||||
C***********************************************************************
|
||||
C***********************************************************************
|
||||
|
||||
SUBROUTINE INIT_COVARIANCES(Ntime,def,R0,R1,R2,R3,R4)
|
||||
IMPLICIT NONE
|
||||
double precision, dimension(:),intent(out) :: R0,R1,R2
|
||||
double precision, dimension(:),intent(out) :: R3,R4
|
||||
integer,intent(in) :: Ntime,def
|
||||
integer :: i
|
||||
open (unit=1, file='Cd0.in',STATUS='unknown')
|
||||
open (unit=2, file='Cd1.in',STATUS='unknown')
|
||||
open (unit=3, file='Cd2.in',STATUS='unknown')
|
||||
|
||||
do i=1,Ntime
|
||||
read(1,*) R0(i)
|
||||
read(2,*) R1(i)
|
||||
read(3,*) R2(i)
|
||||
enddo
|
||||
close(1)
|
||||
close(2)
|
||||
close(3)
|
||||
|
||||
if (abs(def).GT.1) then
|
||||
open (unit=4, file='Cd3.in',STATUS='unknown')
|
||||
open (unit=5, file='Cd4.in',STATUS='unknown')
|
||||
|
||||
do i=1,Ntime
|
||||
read(4,*) R3(i)
|
||||
read(5,*) R4(i)
|
||||
enddo
|
||||
|
||||
close(4)
|
||||
close(5)
|
||||
endif
|
||||
return
|
||||
END SUBROUTINE INIT_COVARIANCES
|
||||
|
||||
C***********************************************************************
|
||||
C***********************************************************************
|
||||
|
||||
C**********************************************************************
|
||||
|
||||
SUBROUTINE COV_INPUT(BIG,tn,ts, R0,R1,R2,R3,R4)
|
||||
IMPLICIT NONE
|
||||
double precision, dimension(:,:),intent(inout) :: BIG
|
||||
double precision, dimension(:),intent(in) :: R0,R1,R2
|
||||
double precision, dimension(:),intent(in) :: R3,R4
|
||||
integer ,intent(in) :: tn,ts
|
||||
integer :: i,j,shft,Ntd1,N !=Ntdc
|
||||
double precision :: tmp
|
||||
! the order of the variables in the covariance matrix
|
||||
! are organized as follows:
|
||||
! For ts>1:
|
||||
! ||X(t2)..X(ts),..X(tn-1)||X''(ts) X'(t1) X'(tn)||X(ts) X(t1) X(tn) X'(ts)||
|
||||
! = [Xt Xd Xc]
|
||||
!
|
||||
! For ts<=1:
|
||||
! ||X(t2)..,..X(tn-1)||X'(t1) X'(tn)||X(t1) X(tn)||
|
||||
! = [Xt Xd Xc]
|
||||
|
||||
! where
|
||||
!
|
||||
! Xt= time points in the indicator function
|
||||
! Xd= derivatives
|
||||
! Xc=variables to condition on
|
||||
|
||||
if (ts.LE.1) THEN
|
||||
Ntd1=tn
|
||||
N=Ntd1+2;
|
||||
shft=0 ! def=1 want only crest period Tc
|
||||
else
|
||||
Ntd1=tn+1
|
||||
N=Ntd1+4
|
||||
shft=1 ! def=2 or 3 want Tc Ac or Tcf, Ac
|
||||
endif
|
||||
|
||||
do i=1,tn-2
|
||||
!cov(Xt)
|
||||
do j=i,tn-2
|
||||
BIG(i,j) = R0(j-i+1) ! cov(X(ti+1),X(tj+1))
|
||||
enddo
|
||||
!cov(Xt,Xc)
|
||||
BIG(i ,Ntd1+1+shft) = R0(i+1) !cov(X(ti+1),X(t1))
|
||||
BIG(tn-1-i ,Ntd1+2+shft) = R0(i+1) !cov(X(t.. ),X(tn))
|
||||
!Cov(Xt,Xd)=cov(X(ti+1),x(tj)
|
||||
BIG(i,Ntd1-1) =-R1(i+1) !cov(X(ti+1),X' (t1))
|
||||
BIG(tn-1-i,Ntd1)= R1(i+1) !cov(X(ti+1),X' (tn))
|
||||
enddo
|
||||
!call echo(big(1:tn,1:tn),tn)
|
||||
!cov(Xd)
|
||||
BIG(Ntd1 ,Ntd1 ) = -R2(1)
|
||||
BIG(Ntd1-1,Ntd1 ) = -R2(tn) !cov(X'(t1),X'(tn))
|
||||
BIG(Ntd1-1,Ntd1-1) = -R2(1)
|
||||
|
||||
!cov(Xc)
|
||||
BIG(Ntd1+1+shft,Ntd1+1+shft) = R0(1) ! cov(X(t1),X (t1))
|
||||
BIG(Ntd1+1+shft,Ntd1+2+shft) = R0(tn) ! cov(X(t1),X (tn))
|
||||
BIG(Ntd1+2+shft,Ntd1+2+shft) = R0(1) ! cov(X(tn),X (tn))
|
||||
!cov(Xd,Xc)
|
||||
BIG(Ntd1 ,Ntd1+1+shft) = R1(tn) !cov(X'(tn),X(t1))
|
||||
BIG(Ntd1 ,Ntd1+2+shft) = 0.d0 !cov(X'(tn),X(tn))
|
||||
BIG(Ntd1-1,Ntd1+1+shft) = 0.d0 !cov(X'(t1),X(t1))
|
||||
BIG(Ntd1-1,Ntd1+2+shft) =-R1(tn) !cov(X'(t1),X(tn))
|
||||
|
||||
|
||||
if (ts.GT.1) then
|
||||
|
||||
!
|
||||
!cov(Xc)
|
||||
BIG(Ntd1+1,Ntd1+1) = R0(1) ! cov(X(ts),X (ts)
|
||||
BIG(Ntd1+1,Ntd1+2) = R0(ts) ! cov(X(ts),X (t1))
|
||||
BIG(Ntd1+1,Ntd1+3) = R0(tn+1-ts) ! cov(X(ts),X (tn))
|
||||
BIG(Ntd1+1,Ntd1+4) = 0.d0 ! cov(X(ts),X'(ts))
|
||||
|
||||
BIG(Ntd1+2,Ntd1+4) = R1(ts) ! cov(X(t1),X'(ts))
|
||||
BIG(Ntd1+3,Ntd1+4) = -R1(tn+1-ts) !cov(X(tn),X'(ts))
|
||||
BIG(Ntd1+4,Ntd1+4) = -R2(1) ! cov(X'(ts),X'(ts))
|
||||
|
||||
!cov(Xd)
|
||||
BIG(Ntd1-2,Ntd1-1) = -R3(ts) !cov(X''(ts),X'(t1))
|
||||
BIG(Ntd1-2,Ntd1-2) = R4(1)
|
||||
BIG(Ntd1-2,Ntd1 ) = R3(tn+1-ts) !cov(X''(ts),X'(tn))
|
||||
!cov(Xd,Xc)
|
||||
BIG(Ntd1 ,Ntd1+4) =-R2(tn+1-ts) !cov(X'(tn),X'(ts))
|
||||
BIG(Ntd1 ,Ntd1+1) = R1(tn+1-ts) !cov(X'(tn),X (ts))
|
||||
|
||||
BIG(Ntd1-1,Ntd1+4) =-R2(ts) !cov(X'(t1),X'(ts))
|
||||
BIG(Ntd1-1,Ntd1+1) =-R1(ts) !cov(X'(t1),X (ts))
|
||||
|
||||
BIG(Ntd1-2,Ntd1+1) = R2(1) !cov(X''(ts),X (ts)
|
||||
BIG(Ntd1-2,Ntd1+2) = R2(ts) !cov(X''(ts),X (t1))
|
||||
BIG(Ntd1-2,Ntd1+3) = R2(tn+1-ts) !cov(X''(ts),X (tn))
|
||||
BIG(Ntd1-2,Ntd1+4) = 0.d0 !cov(X''(ts),X'(ts))
|
||||
!cov(Xt,Xc)
|
||||
do i=1,tn-2
|
||||
j=abs(i+1-ts)
|
||||
BIG(i,Ntd1+1) = R0(j+1) !cov(X(ti+1),X(ts))
|
||||
BIG(i,Ntd1+4) = sign(R1(j+1),R1(j+1)*dble(ts-i-1)) !cov(X(ti+1),X'(ts)) ! check this
|
||||
|
||||
!Cov(Xt,Xd)=cov(X(ti+1),X(ts))
|
||||
BIG(i,Ntd1-2) = R2(j+1) !cov(X(ti+1),X''(ts))
|
||||
enddo
|
||||
endif ! ts>1
|
||||
|
||||
!call echo(big(1:N,1:N),N)
|
||||
! make lower triangular part equal to upper
|
||||
do j=1,N-1
|
||||
do i=j+1,N
|
||||
tmp =BIG(j,i)
|
||||
|
||||
BIG(i,j)=tmp
|
||||
enddo
|
||||
!call echo(big(1:N,1:N),N)
|
||||
|
||||
enddo
|
||||
!call echo(big(1:N,1:N),N)
|
||||
|
||||
C write (*,10) ((BIG(j,i),i=N+1,N+6),j=N+1,N+6)
|
||||
C 10 format(6F8.4)
|
||||
RETURN
|
||||
END SUBROUTINE COV_INPUT
|
||||
|
||||
SUBROUTINE COV_INPUT2(BIG,pt, R0,R1,R2)
|
||||
IMPLICIT NONE
|
||||
double precision, dimension(:,:), intent(out) :: BIG
|
||||
double precision, dimension(:), intent(in) :: R0,R1,R2
|
||||
integer :: pt,i,j
|
||||
! the order of the variables in the covariance matrix
|
||||
! are organized as follows;
|
||||
! X(t2)...X(tn-1) X'(t1) X'(tn) X(t1) X(tn) = [Xt Xd Xc]
|
||||
!
|
||||
! where Xd is the derivatives
|
||||
!
|
||||
! Xt= time points in the indicator function
|
||||
! Xd= derivatives
|
||||
! Xc=variables to condition on
|
||||
|
||||
!cov(Xc)
|
||||
BIG(pt+2,pt+2) = R0(1)
|
||||
BIG(pt+1,pt+1) = R0(1)
|
||||
BIG(pt+1,pt+2) = R0(pt)
|
||||
!cov(Xd)
|
||||
BIG(pt,pt) = -R2(1)
|
||||
BIG(pt-1,pt-1) = -R2(1)
|
||||
BIG(pt-1,pt) = -R2(pt)
|
||||
!cov(Xd,Xc)
|
||||
BIG(pt,pt+2) = 0.d0
|
||||
BIG(pt,pt+1) = R1(pt)
|
||||
BIG(pt-1,pt+2) = -R1(pt)
|
||||
BIG(pt-1,pt+1) = 0.d0
|
||||
|
||||
if (pt.GT.2) then
|
||||
!cov(Xt)
|
||||
do i=1,pt-2
|
||||
do j=i,pt-2
|
||||
BIG(i,j) = R0(j-i+1)
|
||||
enddo
|
||||
enddo
|
||||
!cov(Xt,Xc)
|
||||
do i=1,pt-2
|
||||
BIG(i,pt+1) = R0(i+1)
|
||||
BIG(pt-1-i,pt+2) = R0(i+1)
|
||||
enddo
|
||||
!Cov(Xt,Xd)=cov(X(ti+1),x(tj))
|
||||
do i=1,pt-2
|
||||
BIG(i,pt-1) = -R1(i+1)
|
||||
BIG(pt-1-i,pt)= R1(i+1)
|
||||
enddo
|
||||
endif
|
||||
|
||||
|
||||
! make lower triangular part equal to upper
|
||||
do j=1,pt+1
|
||||
do i=j+1,pt+2
|
||||
BIG(i,j)=BIG(j,i)
|
||||
enddo
|
||||
enddo
|
||||
C write (*,10) ((BIG(j,i),i=N+1,N+6),j=N+1,N+6)
|
||||
C 10 format(6F8.4)
|
||||
RETURN
|
||||
END SUBROUTINE COV_INPUT2
|
||||
|
||||
|
||||
END PROGRAM sp2thpdf
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
@ -1,632 +0,0 @@
|
||||
PROGRAM sp2thpdf
|
||||
!***********************************************************************
|
||||
! This program computes: *
|
||||
! *
|
||||
! density of S_i,Hi,T_i in a gaussian process i.e. *
|
||||
! *
|
||||
! quart wavelength (up-crossing to crest) and crest amplitude *
|
||||
!
|
||||
! def = 1, gives half wave period, Tc (default).
|
||||
! -1, gives half wave period, Tt.
|
||||
! 2, gives half wave period and wave crest amplitude (Tc,Ac).
|
||||
! -2, gives half wave period and wave trough amplitude (Tt,At).
|
||||
! 3, gives crest front period and wave crest amplitude (Tcf,Ac).
|
||||
! -3, gives trough back period and wave trough amplitude (Ttb,At).
|
||||
! 4, gives minimum of crest front/back period and wave crest
|
||||
! amplitude (min(Tcf,Tcb),Ac).
|
||||
! -4, gives minimum of trough front/back period and wave trough
|
||||
! amplitude (min(Ttf,Ttb),At).
|
||||
!***********************************************************************
|
||||
!History:
|
||||
! revised Per A. Brodtkorb 04.04.2000
|
||||
! -
|
||||
! revised Per A. Brodtkorb 23.11.99
|
||||
! - fixed a bug in calculating pdf for def = +/- 4
|
||||
! revised Per A. Brodtkorb 03.11.99
|
||||
! - added def = +/-4
|
||||
! revised Per A. Brodtkorb 23.09.99
|
||||
! - minor changes to covinput
|
||||
! - removed the calculation of the transformation to spec2thpdf.m
|
||||
! by Igor Rychlik
|
||||
|
||||
|
||||
use GLOBALDATA, only : rateLHD,SCIS,NSIMmax,COV,ABSEPS
|
||||
use globalconst
|
||||
use rind
|
||||
IMPLICIT NONE
|
||||
double precision, dimension(:,:),allocatable :: BIG
|
||||
double precision, dimension(:,:),allocatable :: ansr
|
||||
double precision, dimension(: ),allocatable :: ex
|
||||
double precision, dimension(:,:),allocatable :: xc
|
||||
double precision, dimension(: ),allocatable :: fxind,h
|
||||
double precision, dimension(: ),allocatable :: R0,R1,R2,R3,R4
|
||||
double precision ::CC,U,XddInf,XdInf,XtInf
|
||||
double precision, dimension(2,6) :: a_up=0.d0,a_lo=0.d0
|
||||
integer, dimension(6) :: INFIN=2
|
||||
integer , dimension(: ),allocatable :: seed
|
||||
integer ,dimension(7) :: indI
|
||||
integer :: Nx,Nt,Nc,Nd,NI,Mb,Ntd, Ntdc
|
||||
integer :: Nstart,Ntime,tn,ts,speed,ph,def,seed1,seed_size
|
||||
double precision :: dT, EPSOLD ! lag spacing for covariances
|
||||
LOGICAL :: init=.TRUE.
|
||||
! DIGITAL:
|
||||
! f90 -g2 -C -automatic -o ../wave/alpha/sp2thpdf.exe rind44.f sp2thpdf.f
|
||||
! SOLARIS:
|
||||
!f90 -g -O -w3 -Bdynamic -fixed -o ../wave/sol2/sp2thpdf.exe rind44.f sp2thpdf.f
|
||||
! linux:
|
||||
! f90 -gline -Nl126 -C -o ../exec/lnx86/sp2thpdf8.exe intmodule.f rind60.f sp2thpdf.f
|
||||
! f90 -gline -Nl126 -C -o sp2thpdf.exe rind45.f sp2thpdf.f
|
||||
! f90 -gline -Nl126 -C -o ../exec/lnx86/sp2thpdf3.exe adaptmodule.f krbvrcmod.f krobovmod.f rcrudemod.f rind55.f sp2thpdf.f
|
||||
! HP700
|
||||
!f90 -g -C -o ../exec/hp700/sp2thpdf.exe rind45.f sp2thpdf.f
|
||||
!f90 -g -C +check=all +FPVZID -o ../exec/hp700/sp2thpdf.exe rind45.f sp2thpdf.f
|
||||
! f90 +gprof +extend_source +Oall +Odataprefetch +Ofastaccess +Oinfo +Oprocelim -C +check=all -o ../exec/hp700/sp2thpdf.exe rind48.f sp2thpdf.f
|
||||
|
||||
!print *,'enter sp2thpdf'
|
||||
|
||||
CALL INIT_LEVELS(U,def,Ntime,Nstart,speed,SCIS,seed1,
|
||||
& Nx,dT,rateLHD)
|
||||
!print *,'U,def,Ntime,Nstart,NIT,speed,SCIS,seed1,Nx,dT'
|
||||
!print *,U,def,Ntime,Nstart,NIT,speed,SCIS,seed1,Nx,dT
|
||||
|
||||
|
||||
if (SCIS.GT.0) then
|
||||
!allocate(COV(1:Nx))
|
||||
call random_seed(SIZE=seed_size)
|
||||
allocate(seed(seed_size))
|
||||
call random_seed(GET=seed(1:seed_size)) ! get current seed
|
||||
seed(1)=seed1 ! change seed
|
||||
call random_seed(PUT=seed(1:seed_size))
|
||||
deallocate(seed)
|
||||
endif
|
||||
CALL INITDATA(speed)
|
||||
!print *,ntime,speed,u,NIT
|
||||
allocate(R0(1:Ntime+1))
|
||||
allocate(R1(1:Ntime+1))
|
||||
allocate(R2(1:Ntime+1))
|
||||
if (abs(def).GT.1) THEN
|
||||
allocate(h(1:Nx))
|
||||
allocate(R3(1:Ntime+1))
|
||||
allocate(R4(1:Ntime+1))
|
||||
CALL INIT_AMPLITUDES(h,def,Nx)
|
||||
endif
|
||||
CALL INIT_COVARIANCES(Ntime,def,R0,R1,R2,R3,R4)
|
||||
|
||||
!print *,'Nx',Nx
|
||||
|
||||
|
||||
indI(1)=0
|
||||
C ***** The bound 'infinity' is set to 10*sigma *****
|
||||
XdInf=10.d0*SQRT(-R2(1))
|
||||
XtInf=10.d0*SQRT(R0(1))
|
||||
!print *,'XdInf,XtInf'
|
||||
!print *,XdInf,XtInf
|
||||
! normalizing constant
|
||||
CC=TWPI*SQRT(-R0(1)/R2(1))*exp(u*u/(2.d0*R0(1)) )
|
||||
allocate(ansr(1:Ntime,1:Nx))
|
||||
ansr=0.d0
|
||||
allocate(fxind(1:Nx))
|
||||
!fxind=0.d0 this is not needed
|
||||
|
||||
if (abs(def).GT.1) GOTO 200
|
||||
|
||||
NI=4; Nd=2
|
||||
Nc=2; Mb=1
|
||||
Nx=1
|
||||
allocate(BIG(1:Ntime+Nc,1:Ntime+Nc))
|
||||
allocate(xc(1:Nc,1:Nx))
|
||||
allocate(ex(1:Ntime+Nc))
|
||||
ex=0.d0
|
||||
xc(1,1)=u
|
||||
xc(2,1)=u
|
||||
! INFIN = INTEGER, array of integration limits flags: size 1 x Nb (in)
|
||||
! if INFIN(I) < 0, Ith limits are (-infinity, infinity);
|
||||
! if INFIN(I) = 0, Ith limits are (-infinity, Hup(I)];
|
||||
! if INFIN(I) = 1, Ith limits are [Hlo(I), infinity);
|
||||
! if INFIN(I) = 2, Ith limits are [Hlo(I), Hup(I)].
|
||||
|
||||
if (def.GT.0) then
|
||||
INFIN(1:2) = 1
|
||||
INFIN(3) = 0
|
||||
a_up(1,1)= u+XtInf
|
||||
a_lo(1,1)= u
|
||||
a_up(1,2)= XdInf
|
||||
a_lo(1,3)=-XdInf
|
||||
else
|
||||
INFIN(1:2) = 0
|
||||
INFIN(3) = 1
|
||||
a_up(1,1)=u
|
||||
a_lo(1,1)=u-XtInf
|
||||
a_lo(1,2)=-XdInf
|
||||
a_up(1,3)= XdInf
|
||||
endif
|
||||
!print *,'Nstart',Nstart
|
||||
Nstart=MAX(2,Nstart)
|
||||
!print *,'Nstart',Nstart
|
||||
if (ALLOCATED(COV)) then
|
||||
open (unit=11, file='COV.out', STATUS='unknown')
|
||||
write(11,*) 0.d0
|
||||
endif
|
||||
do Ntd=Nstart,Ntime
|
||||
!CALL COV_INPUT2(BIG,Ntd, R0,R1,R2)
|
||||
CALL COV_INPUT(BIG,Ntd,-1,R0,R1,R2,R3,R4) ! positive wave period
|
||||
Nt=Ntd-Nd;
|
||||
indI(2)=Nt;
|
||||
indI(3)=Nt+1;
|
||||
indI(4)=Ntd;
|
||||
Ntdc=Ntd+Nc;
|
||||
! IF (Ntd.GT.5.AND.(INIT)) THEN
|
||||
! INIT=.FALSE.
|
||||
! CALL INITDATA(speed)
|
||||
! ENDIF
|
||||
!if (SCIS.gt.1) Nj=Nt
|
||||
!if (SCIS.gt.0) then
|
||||
! if (SCIS.EQ.2) then
|
||||
! Nj=max(Nt,0)
|
||||
! else
|
||||
! Nj=min(max(Nt-5, 0),0)
|
||||
! endif
|
||||
!endif
|
||||
!Ex=0.d0
|
||||
!CALL echo(BIG(1:Ntdc,1:min(7,Ntdc)),Ntdc)
|
||||
CALL RINDD(fxind,Big(1:Ntdc,1:Ntdc),ex(1:Ntdc),xc,
|
||||
& Nt,indI(1:NI),a_lo(1:Mb,1:NI-1),a_up(1:Mb,1:NI-1),
|
||||
& INFIN(1:NI-1))
|
||||
ansr(Ntd,1)=fxind(1)*CC
|
||||
if (ALLOCATED(COV)) then !SCIS.GT.0
|
||||
write(11,*) COV(1) ! save coefficient of variation
|
||||
endif
|
||||
print *,'Ready: ',Ntd,' of ',Ntime
|
||||
enddo
|
||||
if (ALLOCATED(COV)) then
|
||||
close(11)
|
||||
endif
|
||||
goto 300
|
||||
200 continue
|
||||
XddInf=10.d0*SQRT(R4(1))
|
||||
NI=7; Nd=3
|
||||
Nc=4; Mb=2
|
||||
allocate(BIG(1:Ntime+Nc+1,1:Ntime+Nc+1))
|
||||
ALLOCATE(xc(1:Nc,1:Nx))
|
||||
allocate(ex(1:Ntime+Nc+1))
|
||||
|
||||
ex=0.d0
|
||||
xc(1,1:Nx)=h(1:Nx)
|
||||
xc(2,1:Nx)=u
|
||||
xc(3,1:Nx)=u
|
||||
xc(4,1:Nx)=0.d0
|
||||
|
||||
! INFIN = INTEGER, array of integration limits flags: size 1 x Nb (in)
|
||||
! if INFIN(I) < 0, Ith limits are (-infinity, infinity);
|
||||
! if INFIN(I) = 0, Ith limits are (-infinity, Hup(I)];
|
||||
! if INFIN(I) = 1, Ith limits are [Hlo(I), infinity);
|
||||
! if INFIN(I) = 2, Ith limits are [Hlo(I), Hup(I)].
|
||||
if (def.GT.0) then
|
||||
INFIN(2)=-1
|
||||
INFIN(4)=0
|
||||
INFIN(5)=1
|
||||
INFIN(6)=0
|
||||
a_up(2,1)=1.d0 !*h
|
||||
a_lo(1,1)=u
|
||||
a_up(1,2)=XtInf ! X(ts) is redundant
|
||||
a_lo(1,2)=-Xtinf
|
||||
a_up(2,2)=1.d0 ! *h
|
||||
a_lo(2,2)=1.d0 ! *h
|
||||
a_up(2,3)=1.d0 !*h
|
||||
a_lo(1,3)=u
|
||||
|
||||
a_lo(1,4)=-XddInf
|
||||
a_up(1,5)= XdInf
|
||||
a_lo(1,6)=-XdInf
|
||||
else !def<0
|
||||
INFIN(2)=-1
|
||||
INFIN(4)=1
|
||||
INFIN(5)=0
|
||||
INFIN(6)=1
|
||||
a_up(1,1)=u
|
||||
a_lo(2,1)=1.d0 !*h
|
||||
a_up(1,2)=XtInf ! X(ts) is redundant
|
||||
a_lo(1,2)=-Xtinf
|
||||
a_up(2,2)=1.d0 ! *h
|
||||
a_lo(2,2)=1.d0 ! *h
|
||||
a_up(1,3)=u
|
||||
a_lo(2,3)=1.d0 !*h
|
||||
a_up(1,4)=XddInf
|
||||
a_lo(1,5)=-XdInf
|
||||
a_up(1,6)=XdInf
|
||||
endif
|
||||
EPSOLD=ABSEPS
|
||||
Nstart=MAX(Nstart,3)
|
||||
do tn=Nstart,Ntime,1
|
||||
Ntd=tn+1
|
||||
Nt=Ntd-Nd
|
||||
Ntdc=Ntd+Nc
|
||||
indI(4)=Nt
|
||||
indI(5)=Nt+1
|
||||
indI(6)=Nt+2
|
||||
indI(7)=Ntd
|
||||
! IF (Ntd.GT.5.AND.INIT) THEN
|
||||
! INIT=.FALSE.
|
||||
! CALL INITDATA(speed)
|
||||
! ENDIF
|
||||
!if (SCIS.gt.1) Nj=Nt
|
||||
!if (SCIS.gt.0) then
|
||||
! if (SCIS.EQ.2) then
|
||||
! Nj=max(Nt,0)
|
||||
! else
|
||||
! Nj=min(max(Nt-5, 0),0)
|
||||
! endif
|
||||
!endif
|
||||
ABSEPS=MIN(SQRT(DBLE(tn))*EPSOLD*0.5D0,0.1D0)
|
||||
do ts=2,FLOOR(DBLE(tn+1)/2.d0)
|
||||
!print *,'ts,tn' ,ts,tn
|
||||
CALL COV_INPUT(Big(1:Ntdc,1:Ntdc),tn,ts,R0,R1,R2,R3,R4) ! positive wave period
|
||||
indI(2)=ts-2
|
||||
indI(3)=ts-1
|
||||
!CALL echo(BIG(1:Ntdc,1:min(7,Ntdc)),Ntdc)
|
||||
!print *,'sp call rind'
|
||||
CALL RINDD(fxind,Big(1:Ntdc,1:Ntdc),ex(1:Ntdc),xc,
|
||||
& Nt,indI(1:NI),a_lo(1:Mb,1:NI-1),a_up(1:Mb,1:NI-1),
|
||||
& INFIN(1:NI-1))
|
||||
!CALL echo(BIG(1:Ntdc,1:min(7,Ntdc)),Ntdc)
|
||||
!print *,'sp rind finished',fxind
|
||||
!goto 900
|
||||
SELECT CASE (ABS(def))
|
||||
CASE (:2)
|
||||
! 2, gives half wave period and wave crest amplitude (Tc,Ac).
|
||||
! -2, gives half wave period and wave trough amplitude (Tt,At).
|
||||
if (ts .EQ.tn-ts+1) then
|
||||
ansr(tn,1:Nx)=ansr(tn,1:Nx)+fxind*CC*dt
|
||||
else
|
||||
ansr(tn,1:Nx)=ansr(tn,1:Nx)+fxind*CC*2.d0*dt
|
||||
endif
|
||||
CASE (3)
|
||||
! 3, gives crest front period and wave crest amplitude (Tcf,Ac).
|
||||
! -3, gives trough back period and wave trough amplitude (Ttb,At).
|
||||
ansr(ts,1:Nx)=ansr(ts,1:Nx)+fxind*CC*dT
|
||||
if ((ts.LT.tn-ts+1)) THEN
|
||||
ansr(tn-ts+1,1:Nx)=ansr(tn-ts+1,1:Nx)+fxind*CC*dT ! exploiting the symmetry
|
||||
endif
|
||||
CASE (4:)
|
||||
! 4, gives minimum of crest front/back period and wave crest amplitude (min(Tcf,Tcb),Ac).
|
||||
! -4, gives minimum of trough front/back period and wave trough amplitude (min(Ttf,Ttb),At).
|
||||
if (ts .EQ.tn-ts+1) then
|
||||
ansr(ts,1:Nx)=ansr(ts,1:Nx)+fxind*CC*dt
|
||||
else
|
||||
ansr(ts,1:Nx)=ansr(ts,1:Nx)+fxind*CC*2.0*dt
|
||||
endif
|
||||
end select
|
||||
enddo ! ts
|
||||
print *,'Ready: ',tn,' of ',Ntime, ' ABSEPS = ', ABSEPS
|
||||
|
||||
enddo !tn
|
||||
!print *,'ansr',ansr
|
||||
300 open (unit=11, file='dens.out', STATUS='unknown')
|
||||
!print *, ansr
|
||||
do ts=1,Ntime
|
||||
do ph=1,Nx
|
||||
write(11,*) ansr(ts,ph)
|
||||
! write(11,111) ansr(ts,ph)
|
||||
enddo
|
||||
enddo
|
||||
!111 FORMAT(2x,F12.8)
|
||||
close(11)
|
||||
900 deallocate(big)
|
||||
deallocate(fxind)
|
||||
deallocate(ansr)
|
||||
deallocate(xc)
|
||||
deallocate(ex)
|
||||
deallocate(R0)
|
||||
deallocate(R1)
|
||||
deallocate(R2)
|
||||
if (allocated(COV) ) then
|
||||
deallocate(COV)
|
||||
endif
|
||||
if (allocated(R3)) then
|
||||
deallocate(R3)
|
||||
deallocate(R4)
|
||||
deallocate(h)
|
||||
ENDIF
|
||||
stop
|
||||
!return
|
||||
|
||||
CONTAINS
|
||||
|
||||
|
||||
|
||||
SUBROUTINE INIT_LEVELS
|
||||
& (U,def,Ntime,Nstart,speed,SCIS,seed1,Nx,dT,rateLHD)
|
||||
IMPLICIT NONE
|
||||
integer, intent(out):: def,Ntime,Nstart,speed,Nx,SCIS,seed1,
|
||||
& rateLHD
|
||||
double precision ,intent(out) :: U,dT
|
||||
double precision :: XSPLT
|
||||
integer :: NIT
|
||||
|
||||
|
||||
OPEN(UNIT=14,FILE='reflev.in',STATUS= 'UNKNOWN')
|
||||
READ (14,*) U
|
||||
READ (14,*) def
|
||||
READ (14,*) Ntime
|
||||
READ (14,*) Nstart
|
||||
READ (14,*) NIT
|
||||
READ (14,*) speed
|
||||
READ (14,*) SCIS
|
||||
READ (14,*) seed1
|
||||
READ (14,*) Nx
|
||||
READ (14,*) dT
|
||||
READ (14,*) rateLHD
|
||||
READ (14,*) XSPLT
|
||||
if (abs(def).GT.1) then
|
||||
|
||||
if (Ntime.lt.3) then
|
||||
print *,'The number of wavelength points is too small, stop'
|
||||
stop
|
||||
end if
|
||||
else
|
||||
Nx=1
|
||||
if (Ntime.lt.2) then
|
||||
print *,'The number of wavelength points is too small, stop'
|
||||
stop
|
||||
end if
|
||||
endif
|
||||
CLOSE(UNIT=14)
|
||||
|
||||
RETURN
|
||||
END SUBROUTINE INIT_LEVELS
|
||||
|
||||
C******************************************************
|
||||
SUBROUTINE INIT_AMPLITUDES(h,def,Nx)
|
||||
IMPLICIT NONE
|
||||
double precision, dimension(:), intent(out) :: h
|
||||
integer, intent(in) :: def
|
||||
integer, intent(in) :: Nx
|
||||
integer :: ix
|
||||
|
||||
|
||||
OPEN(UNIT=4,FILE='h.in',STATUS= 'UNKNOWN')
|
||||
|
||||
C
|
||||
C Reading in amplitudes
|
||||
C
|
||||
do ix=1,Nx
|
||||
READ (4,*) H(ix)
|
||||
enddo
|
||||
CLOSE(UNIT=4)
|
||||
!if (def.LT.0) THEN
|
||||
! H=-H
|
||||
!endif
|
||||
|
||||
RETURN
|
||||
END SUBROUTINE INIT_AMPLITUDES
|
||||
|
||||
C**************************************************
|
||||
|
||||
C***********************************************************************
|
||||
C***********************************************************************
|
||||
|
||||
SUBROUTINE INIT_COVARIANCES(Ntime,def,R0,R1,R2,R3,R4)
|
||||
IMPLICIT NONE
|
||||
double precision, dimension(:),intent(out) :: R0,R1,R2
|
||||
double precision, dimension(:),intent(out) :: R3,R4
|
||||
integer,intent(in) :: Ntime,def
|
||||
integer :: i
|
||||
open (unit=1, file='Cd0.in',STATUS='unknown')
|
||||
open (unit=2, file='Cd1.in',STATUS='unknown')
|
||||
open (unit=3, file='Cd2.in',STATUS='unknown')
|
||||
|
||||
do i=1,Ntime
|
||||
read(1,*) R0(i)
|
||||
read(2,*) R1(i)
|
||||
read(3,*) R2(i)
|
||||
enddo
|
||||
close(1)
|
||||
close(2)
|
||||
close(3)
|
||||
|
||||
if (abs(def).GT.1) then
|
||||
open (unit=4, file='Cd3.in',STATUS='unknown')
|
||||
open (unit=5, file='Cd4.in',STATUS='unknown')
|
||||
|
||||
do i=1,Ntime
|
||||
read(4,*) R3(i)
|
||||
read(5,*) R4(i)
|
||||
enddo
|
||||
|
||||
close(4)
|
||||
close(5)
|
||||
endif
|
||||
return
|
||||
END SUBROUTINE INIT_COVARIANCES
|
||||
|
||||
C***********************************************************************
|
||||
C***********************************************************************
|
||||
|
||||
C**********************************************************************
|
||||
|
||||
SUBROUTINE COV_INPUT(BIG,tn,ts, R0,R1,R2,R3,R4)
|
||||
IMPLICIT NONE
|
||||
double precision, dimension(:,:),intent(inout) :: BIG
|
||||
double precision, dimension(:),intent(in) :: R0,R1,R2
|
||||
double precision, dimension(:),intent(in) :: R3,R4
|
||||
integer ,intent(in) :: tn,ts
|
||||
integer :: i,j,shft,Ntd1,N !=Ntdc
|
||||
! the order of the variables in the covariance matrix
|
||||
! are organized as follows:
|
||||
! For ts>1:
|
||||
! ||X(t2)..X(ts),..X(tn-1)||X''(ts) X'(t1) X'(tn)||X(ts) X(t1) X(tn) X'(ts)||
|
||||
! = [Xt Xd Xc]
|
||||
!
|
||||
! For ts<=1:
|
||||
! ||X(t2)..,..X(tn-1)||X'(t1) X'(tn)||X(t1) X(tn)||
|
||||
! = [Xt Xd Xc]
|
||||
|
||||
! where
|
||||
!
|
||||
! Xt= time points in the indicator function
|
||||
! Xd= derivatives
|
||||
! Xc=variables to condition on
|
||||
|
||||
if (ts.LE.1) THEN
|
||||
Ntd1=tn
|
||||
N=Ntd1+2;
|
||||
shft=0 ! def=1 want only crest period Tc
|
||||
else
|
||||
Ntd1=tn+1
|
||||
N=Ntd1+4
|
||||
shft=1 ! def=2 or 3 want Tc Ac or Tcf, Ac
|
||||
endif
|
||||
|
||||
do i=1,tn-2
|
||||
!cov(Xt)
|
||||
do j=i,tn-2
|
||||
BIG(i,j) = R0(j-i+1) ! cov(X(ti+1),X(tj+1))
|
||||
enddo
|
||||
!cov(Xt,Xc)
|
||||
BIG(i ,Ntd1+1+shft) = R0(i+1) !cov(X(ti+1),X(t1))
|
||||
BIG(tn-1-i ,Ntd1+2+shft) = R0(i+1) !cov(X(t.. ),X(tn))
|
||||
!Cov(Xt,Xd)=cov(X(ti+1),x(tj)
|
||||
BIG(i,Ntd1-1) =-R1(i+1) !cov(X(ti+1),X' (t1))
|
||||
BIG(tn-1-i,Ntd1)= R1(i+1) !cov(X(ti+1),X' (tn))
|
||||
enddo
|
||||
!call echo(big(1:tn,1:tn),tn)
|
||||
!cov(Xd)
|
||||
BIG(Ntd1 ,Ntd1 ) = -R2(1)
|
||||
BIG(Ntd1-1,Ntd1 ) = -R2(tn) !cov(X'(t1),X'(tn))
|
||||
BIG(Ntd1-1,Ntd1-1) = -R2(1)
|
||||
|
||||
!cov(Xc)
|
||||
BIG(Ntd1+1+shft,Ntd1+1+shft) = R0(1) ! cov(X(t1),X (t1))
|
||||
BIG(Ntd1+1+shft,Ntd1+2+shft) = R0(tn) ! cov(X(t1),X (tn))
|
||||
BIG(Ntd1+2+shft,Ntd1+2+shft) = R0(1) ! cov(X(tn),X (tn))
|
||||
!cov(Xd,Xc)
|
||||
BIG(Ntd1 ,Ntd1+1+shft) = R1(tn) !cov(X'(tn),X(t1))
|
||||
BIG(Ntd1 ,Ntd1+2+shft) = 0.d0 !cov(X'(tn),X(tn))
|
||||
BIG(Ntd1-1,Ntd1+1+shft) = 0.d0 !cov(X'(t1),X(t1))
|
||||
BIG(Ntd1-1,Ntd1+2+shft) =-R1(tn) !cov(X'(t1),X(tn))
|
||||
|
||||
|
||||
if (ts.GT.1) then
|
||||
|
||||
!
|
||||
!cov(Xc)
|
||||
BIG(Ntd1+1,Ntd1+1) = R0(1) ! cov(X(ts),X (ts)
|
||||
BIG(Ntd1+1,Ntd1+2) = R0(ts) ! cov(X(ts),X (t1))
|
||||
BIG(Ntd1+1,Ntd1+3) = R0(tn+1-ts) ! cov(X(ts),X (tn))
|
||||
BIG(Ntd1+1,Ntd1+4) = 0.d0 ! cov(X(ts),X'(ts))
|
||||
|
||||
BIG(Ntd1+2,Ntd1+4) = R1(ts) ! cov(X(t1),X'(ts))
|
||||
BIG(Ntd1+3,Ntd1+4) = -R1(tn+1-ts) !cov(X(tn),X'(ts))
|
||||
BIG(Ntd1+4,Ntd1+4) = -R2(1) ! cov(X'(ts),X'(ts))
|
||||
|
||||
!cov(Xd)
|
||||
BIG(Ntd1-2,Ntd1-1) = -R3(ts) !cov(X''(ts),X'(t1))
|
||||
BIG(Ntd1-2,Ntd1-2) = R4(1)
|
||||
BIG(Ntd1-2,Ntd1 ) = R3(tn+1-ts) !cov(X''(ts),X'(tn))
|
||||
!cov(Xd,Xc)
|
||||
BIG(Ntd1 ,Ntd1+4) =-R2(tn+1-ts) !cov(X'(tn),X'(ts))
|
||||
BIG(Ntd1 ,Ntd1+1) = R1(tn+1-ts) !cov(X'(tn),X (ts))
|
||||
|
||||
BIG(Ntd1-1,Ntd1+4) =-R2(ts) !cov(X'(t1),X'(ts))
|
||||
BIG(Ntd1-1,Ntd1+1) =-R1(ts) !cov(X'(t1),X (ts))
|
||||
|
||||
BIG(Ntd1-2,Ntd1+1) = R2(1) !cov(X''(ts),X (ts)
|
||||
BIG(Ntd1-2,Ntd1+2) = R2(ts) !cov(X''(ts),X (t1))
|
||||
BIG(Ntd1-2,Ntd1+3) = R2(tn+1-ts) !cov(X''(ts),X (tn))
|
||||
BIG(Ntd1-2,Ntd1+4) = 0.d0 !cov(X''(ts),X'(ts))
|
||||
!cov(Xt,Xc)
|
||||
do i=1,tn-2
|
||||
j=abs(i+1-ts)
|
||||
BIG(i,Ntd1+1) = R0(j+1) !cov(X(ti+1),X(ts))
|
||||
BIG(i,Ntd1+4) = sign(R1(j+1),R1(j+1)*dble(ts-i-1)) !cov(X(ti+1),X'(ts)) ! check this
|
||||
|
||||
!Cov(Xt,Xd)=cov(X(ti+1),X(ts))
|
||||
BIG(i,Ntd1-2) = R2(j+1) !cov(X(ti+1),X''(ts))
|
||||
enddo
|
||||
endif ! ts>1
|
||||
|
||||
!call echo(big(1:N,1:N),N)
|
||||
! make lower triangular part equal to upper
|
||||
do j=1,N-1
|
||||
do i=j+1,N
|
||||
BIG(i,j) =BIG(j,i)
|
||||
enddo
|
||||
!call echo(big(1:N,1:N),N)
|
||||
enddo
|
||||
!call echo(big(1:N,1:N),N)
|
||||
|
||||
C write (*,10) ((BIG(j,i),i=N+1,N+6),j=N+1,N+6)
|
||||
C 10 format(6F8.4)
|
||||
RETURN
|
||||
END SUBROUTINE COV_INPUT
|
||||
|
||||
SUBROUTINE COV_INPUT2(BIG,pt, R0,R1,R2)
|
||||
IMPLICIT NONE
|
||||
double precision, dimension(:,:), intent(out) :: BIG
|
||||
double precision, dimension(:), intent(in) :: R0,R1,R2
|
||||
integer :: pt,i,j
|
||||
! the order of the variables in the covariance matrix
|
||||
! are organized as follows;
|
||||
! X(t2)...X(tn-1) X'(t1) X'(tn) X(t1) X(tn) = [Xt Xd Xc]
|
||||
!
|
||||
! where Xd is the derivatives
|
||||
!
|
||||
! Xt= time points in the indicator function
|
||||
! Xd= derivatives
|
||||
! Xc=variables to condition on
|
||||
|
||||
!cov(Xc)
|
||||
BIG(pt+2,pt+2) = R0(1)
|
||||
BIG(pt+1,pt+1) = R0(1)
|
||||
BIG(pt+1,pt+2) = R0(pt)
|
||||
!cov(Xd)
|
||||
BIG(pt,pt) = -R2(1)
|
||||
BIG(pt-1,pt-1) = -R2(1)
|
||||
BIG(pt-1,pt) = -R2(pt)
|
||||
!cov(Xd,Xc)
|
||||
BIG(pt,pt+2) = 0.d0
|
||||
BIG(pt,pt+1) = R1(pt)
|
||||
BIG(pt-1,pt+2) = -R1(pt)
|
||||
BIG(pt-1,pt+1) = 0.d0
|
||||
|
||||
if (pt.GT.2) then
|
||||
!cov(Xt)
|
||||
do i=1,pt-2
|
||||
do j=i,pt-2
|
||||
BIG(i,j) = R0(j-i+1)
|
||||
enddo
|
||||
enddo
|
||||
!cov(Xt,Xc)
|
||||
do i=1,pt-2
|
||||
BIG(i,pt+1) = R0(i+1)
|
||||
BIG(pt-1-i,pt+2) = R0(i+1)
|
||||
enddo
|
||||
!Cov(Xt,Xd)=cov(X(ti+1),x(tj))
|
||||
do i=1,pt-2
|
||||
BIG(i,pt-1) = -R1(i+1)
|
||||
BIG(pt-1-i,pt)= R1(i+1)
|
||||
enddo
|
||||
endif
|
||||
|
||||
|
||||
! make lower triangular part equal to upper
|
||||
do j=1,pt+1
|
||||
do i=j+1,pt+2
|
||||
BIG(i,j)=BIG(j,i)
|
||||
enddo
|
||||
enddo
|
||||
C write (*,10) ((BIG(j,i),i=N+1,N+6),j=N+1,N+6)
|
||||
C 10 format(6F8.4)
|
||||
RETURN
|
||||
END SUBROUTINE COV_INPUT2
|
||||
|
||||
END PROGRAM sp2thpdf
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
@ -1,505 +0,0 @@
|
||||
PROGRAM sp2tthpdf
|
||||
C***********************************************************************
|
||||
C This program computes upper and lower bounds for the: *
|
||||
C *
|
||||
C density of T= T_1+T_2 in a gaussian process i.e. *
|
||||
C *
|
||||
C wavelengthes for crests <h1 and troughs >h2 *
|
||||
C *
|
||||
C Sylvie and Igor 7 dec. 1999 *
|
||||
C***********************************************************************
|
||||
use GLOBALDATA, only : Nt,Nj,Nd,Nc,Ntd,Ntdc,NI,Mb,
|
||||
& NIT,Nx,TWOPI,XSPLT,SCIS,NSIMmax,COV
|
||||
use rind
|
||||
IMPLICIT NONE
|
||||
double precision, dimension(:,:),allocatable :: BIG
|
||||
double precision, dimension(:,:),allocatable :: ansrup
|
||||
double precision, dimension(:,:),allocatable :: ansrlo
|
||||
double precision, dimension(: ),allocatable :: ex,CY1,CY2
|
||||
double precision, dimension(:,:),allocatable :: xc
|
||||
double precision, dimension(:,:),allocatable ::fxind
|
||||
double precision, dimension(: ),allocatable :: h1,h2
|
||||
double precision, dimension(: ),allocatable :: hh1,hh2
|
||||
double precision, dimension(: ),allocatable :: R0,R1,R2
|
||||
double precision ::CC,U,XddInf,XdInf,XtInf
|
||||
double precision, dimension(:,:),allocatable :: a_up,a_lo
|
||||
integer , dimension(: ),allocatable :: seed
|
||||
integer ,dimension(7) :: indI
|
||||
integer :: Ntime,N0,tn,ts,speed,ph,seed1,seed_size,Nx1,Nx2
|
||||
integer :: icy,icy2
|
||||
double precision :: ds,dT ! lag spacing for covariances
|
||||
! DIGITAL:
|
||||
! f90 -g2 -C -automatic -o ~/WAT/V4/sp2tthpdf.exe rind48.f sp2tthpdf.f
|
||||
! SOLARIS:
|
||||
!f90 -g -O -w3 -Bdynamic -fixed -o ../sp2tthpdf.exe rind48.f sp2tthpdf.f
|
||||
|
||||
!print *,'enter sp2thpdf'
|
||||
CALL INIT_LEVELS(U,Ntime,N0,NIT,speed,SCIS,seed1,Nx1,Nx2,dT)
|
||||
|
||||
!print *,'U,Ntime,NIT,speed,SCIS,seed1,Nx,dT'
|
||||
!print *,U,Ntime,NIT,speed,SCIS,seed1,Nx,dT
|
||||
!Nx1=1
|
||||
!Nx2=1
|
||||
|
||||
Nx=Nx1*Nx2
|
||||
!print *,'NN',Nx1,Nx2,Nx
|
||||
|
||||
|
||||
!XSPLT=1.5d0
|
||||
if (SCIS.GT.0) then
|
||||
allocate(COV(1:Nx))
|
||||
call random_seed(SIZE=seed_size)
|
||||
allocate(seed(seed_size))
|
||||
call random_seed(GET=seed(1:seed_size)) ! get current seed
|
||||
seed(1)=seed1 ! change seed
|
||||
call random_seed(PUT=seed(1:seed_size))
|
||||
deallocate(seed)
|
||||
endif
|
||||
CALL INITDATA(speed)
|
||||
!print *,ntime,speed,u,NIT
|
||||
allocate(R0(1:Ntime+1))
|
||||
allocate(R1(1:Ntime+1))
|
||||
allocate(R2(1:Ntime+1))
|
||||
|
||||
allocate(h1(1:Nx1))
|
||||
allocate(h2(1:Nx2))
|
||||
CALL INIT_AMPLITUDES(h1,Nx1,h2,Nx2)
|
||||
CALL INIT_COVARIANCES(Ntime,R0,R1,R2)
|
||||
|
||||
|
||||
allocate(hh1(1:Nx))
|
||||
allocate(hh2(1:Nx))
|
||||
!h transformation
|
||||
do icy=1,Nx1
|
||||
do icy2=1,Nx2
|
||||
hh1((icy-1)*Nx2+icy2)=h1(icy);
|
||||
hh2((icy-1)*Nx2+icy2)=h2(icy2);
|
||||
enddo
|
||||
enddo
|
||||
|
||||
Nj=0
|
||||
indI(1)=0
|
||||
|
||||
C ***** The bound 'infinity' is set to 10*sigma *****
|
||||
XdInf=10.d0*SQRT(-R2(1))
|
||||
XtInf=10.d0*SQRT(R0(1))
|
||||
!h1(1)=XtInf
|
||||
!h2(1)=XtInf
|
||||
! normalizing constant
|
||||
CC=TWOPI*SQRT(-R0(1)/R2(1))*exp(u*u/(2.d0*R0(1)) )
|
||||
allocate(CY1(1:Nx))
|
||||
allocate(CY2(1:Nx))
|
||||
do icy=1,Nx
|
||||
CY1(icy)=exp(-0.5*hh1(icy)*hh1(icy)/100)/(10*sqrt(twopi))
|
||||
CY2(icy)=exp(-0.5*hh2(icy)*hh2(icy)/100)/(10*sqrt(twopi))
|
||||
enddo
|
||||
!print *,CY1
|
||||
allocate(ansrup(1:Ntime,1:Nx))
|
||||
allocate(ansrlo(1:Ntime,1:Nx))
|
||||
ansrup=0.d0
|
||||
ansrlo=0.d0
|
||||
allocate(fxind(1:Nx,1:2))
|
||||
!fxind=0.d0 this is not needed
|
||||
|
||||
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
|
||||
! Y={X(t2)..,X(ts),..X(tn-1)||X'(ts) X'(t1) X'(tn)||Y1 Y2 X(ts) X(t1) X(tn)} !!
|
||||
! = [Xt Xd Xc] !!
|
||||
! !!
|
||||
! Nt=tn-2, Nd=3, Nc=2+3 !!
|
||||
! !!
|
||||
! Xt= contains Nt time points in the indicator function !!
|
||||
! Xd= " Nd derivatives !!
|
||||
! Xc= " Nc variables to condition on !!
|
||||
! (Y1,Y2) dummy variables ind. of all other v. inputing h1,h2 into rindd !!
|
||||
! !!
|
||||
! There are 6 ( NI=7) regions with constant bariers: !!
|
||||
! (indI(1)=0); for i\in (indI(1),indI(2)] u<Y(i)<h1 !!
|
||||
! (indI(2)=ts-2); for i\in (indI(2),indI(2)], inf<Y(i)<inf (no restr.) !!
|
||||
! (indI(3)=ts-1); for i\in (indI(3),indI(4)], h2 <Y(i)<u !!
|
||||
! (indI(4)=Nt) ; for i\in (indI(4),indI(5)], Y(i)<0 (deriv. X'(ts)) !!
|
||||
! (indI(5)=Nt+1); for i\in (indI(5),indI(6)], Y(i)>0 (deriv. X'(t1)) !!
|
||||
! (indI(6)=Nt+2); for i\in (indI(6),indI(7)], Y(i)>0 (deriv. X'(tn)) !!
|
||||
! (indI(7)=Nt+3); NI=7. !!
|
||||
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
|
||||
|
||||
|
||||
NI=7; Nd=3
|
||||
Nc=5; Mb=3
|
||||
allocate(a_up(1:Mb,1:(NI-1)))
|
||||
allocate(a_lo(1:Mb,1:(NI-1)))
|
||||
a_up=0.d0
|
||||
a_lo=0.d0
|
||||
allocate(BIG(1:(Ntime+Nc+1),1:(Ntime+Nc+1)))
|
||||
ALLOCATE(xc(1:Nc,1:Nx))
|
||||
allocate(ex(1:(Ntime+Nc+1)))
|
||||
!print *,size(ex),Ntime
|
||||
ex=0.d0
|
||||
!print *,size(ex),ex
|
||||
xc(1,1:Nx)=hh1(1:Nx)
|
||||
xc(2,1:Nx)=hh2(1:Nx)
|
||||
xc(3,1:Nx)=u
|
||||
xc(4,1:Nx)=u
|
||||
xc(5,1:Nx)=u
|
||||
! upp- down- upp-crossings at t1,ts,tn
|
||||
|
||||
a_lo(1,1)=u
|
||||
a_up(1,2)=XtInf ! X(ts) is redundant
|
||||
a_lo(1,2)=-Xtinf
|
||||
a_up(1,3)=u
|
||||
|
||||
|
||||
a_lo(1,4)=-XdInf
|
||||
a_up(1,5)= XdInf
|
||||
a_up(1,6)= XdInf
|
||||
|
||||
a_up(2,1)=1.d0
|
||||
a_lo(3,3)=1.d0 !signe a voir!!!!!!
|
||||
! print *,a_up
|
||||
! print *,a_lo
|
||||
do tn=N0,Ntime,1
|
||||
! do tn=Ntime,Ntime,1
|
||||
Ntd=tn+1
|
||||
Nt=Ntd-Nd
|
||||
Ntdc=Ntd+Nc
|
||||
indI(4)=Nt
|
||||
indI(5)=Nt+1
|
||||
indI(6)=Nt+2
|
||||
indI(7)=Ntd
|
||||
if (SCIS.gt.0) then
|
||||
if (SCIS.EQ.2) then
|
||||
Nj=max(Nt,0)
|
||||
else
|
||||
Nj=min(max(Nt-5, 0),0)
|
||||
endif
|
||||
endif
|
||||
do ts=3,tn-2
|
||||
!print *,'ts,tn' ,ts,tn,Ntdc
|
||||
CALL COV_INPUT(Big(1:Ntdc,1:Ntdc),tn,ts,R0,R1,R2)!positive wave period
|
||||
indI(2)=ts-2
|
||||
indI(3)=ts-1
|
||||
|
||||
|
||||
CALL RINDD(fxind,Big(1:Ntdc,1:Ntdc),ex(1:Ntdc),
|
||||
& xc,indI,a_lo,a_up)
|
||||
|
||||
ds=dt
|
||||
do icy=1,Nx
|
||||
! ansr(tn,:)=ansr(tn,:)+fxind*CC*ds./(CY1.*CY2)
|
||||
ansrup(tn,icy)=ansrup(tn,icy)+fxind(icy,1)*CC*ds
|
||||
& /(CY1(icy)*CY2(icy))
|
||||
ansrlo(tn,icy)=ansrlo(tn,icy)+fxind(icy,2)*CC*ds
|
||||
& /(CY1(icy)*CY2(icy))
|
||||
enddo
|
||||
enddo ! ts
|
||||
print *,'Ready: ',tn,' of ',Ntime
|
||||
|
||||
enddo !tn
|
||||
|
||||
300 open (unit=11, file='dens.out', STATUS='unknown')
|
||||
|
||||
do ts=1,Ntime
|
||||
do ph=1,Nx
|
||||
write(11,*) ansrup(ts,ph),ansrlo(ts,ph)!,hh1(ph),hh2(ph)
|
||||
! write(11,111) ansrup(ts,ph),ansrlo(ts,ph)
|
||||
|
||||
enddo
|
||||
enddo
|
||||
!111 FORMAT(2x,F12.8)
|
||||
close(11)
|
||||
900 deallocate(big)
|
||||
deallocate(fxind)
|
||||
deallocate(ansrup)
|
||||
deallocate(ansrlo)
|
||||
deallocate(xc)
|
||||
deallocate(ex)
|
||||
deallocate(R0)
|
||||
deallocate(R1)
|
||||
deallocate(R2)
|
||||
if (allocated(COV) ) then
|
||||
deallocate(COV)
|
||||
endif
|
||||
deallocate(h1)
|
||||
deallocate(h2)
|
||||
deallocate(hh1)
|
||||
deallocate(hh2)
|
||||
deallocate(a_up)
|
||||
deallocate(a_lo)
|
||||
stop
|
||||
!return
|
||||
|
||||
CONTAINS
|
||||
|
||||
|
||||
|
||||
SUBROUTINE INIT_LEVELS
|
||||
& (U,Ntime,N0,NIT,speed,SCIS,seed1,Nx1,Nx2,dT)
|
||||
IMPLICIT NONE
|
||||
integer, intent(out):: Ntime,N0,NIT,speed,Nx1,Nx2,SCIS,seed1
|
||||
double precision ,intent(out) :: U,dT
|
||||
|
||||
|
||||
OPEN(UNIT=14,FILE='reflev.in',STATUS= 'UNKNOWN')
|
||||
READ (14,*) U
|
||||
READ (14,*) Ntime
|
||||
READ (14,*) N0
|
||||
READ (14,*) NIT
|
||||
READ (14,*) speed
|
||||
READ (14,*) SCIS
|
||||
READ (14,*) seed1
|
||||
|
||||
|
||||
READ (14,*) Nx1,Nx2
|
||||
READ (14,*) dT
|
||||
if (Ntime.lt.3) then
|
||||
print *,'The number of wavelength points is too small, stop'
|
||||
stop
|
||||
end if
|
||||
|
||||
CLOSE(UNIT=14)
|
||||
|
||||
RETURN
|
||||
END SUBROUTINE INIT_LEVELS
|
||||
|
||||
C******************************************************
|
||||
SUBROUTINE INIT_AMPLITUDES(h1,Nx1,h2,Nx2)
|
||||
IMPLICIT NONE
|
||||
double precision, dimension(:), intent(out) :: h1,h2
|
||||
integer, intent(in) :: Nx1,Nx2
|
||||
integer :: ix
|
||||
|
||||
|
||||
OPEN(UNIT=4,FILE='h.in',STATUS= 'UNKNOWN')
|
||||
|
||||
C
|
||||
C Reading in amplitudes
|
||||
C
|
||||
do ix=1,Nx1
|
||||
READ (4,*) H1(ix)
|
||||
enddo
|
||||
do ix=1,Nx2
|
||||
READ (4,*) H2(ix)
|
||||
enddo
|
||||
CLOSE(UNIT=4)
|
||||
|
||||
RETURN
|
||||
END SUBROUTINE INIT_AMPLITUDES
|
||||
|
||||
C**************************************************
|
||||
|
||||
C***********************************************************************
|
||||
C***********************************************************************
|
||||
|
||||
SUBROUTINE INIT_COVARIANCES(Ntime,R0,R1,R2)
|
||||
IMPLICIT NONE
|
||||
double precision, dimension(:),intent(out) :: R0,R1,R2
|
||||
integer,intent(in) :: Ntime
|
||||
integer :: i
|
||||
open (unit=1, file='Cd0.in',STATUS='unknown')
|
||||
open (unit=2, file='Cd1.in',STATUS='unknown')
|
||||
open (unit=3, file='Cd2.in',STATUS='unknown')
|
||||
|
||||
do i=1,Ntime
|
||||
read(1,*) R0(i)
|
||||
read(2,*) R1(i)
|
||||
read(3,*) R2(i)
|
||||
enddo
|
||||
close(1)
|
||||
close(2)
|
||||
close(3)
|
||||
|
||||
return
|
||||
END SUBROUTINE INIT_COVARIANCES
|
||||
|
||||
C***********************************************************************
|
||||
C***********************************************************************
|
||||
|
||||
C**********************************************************************
|
||||
|
||||
SUBROUTINE COV_INPUT(BIG,tn,ts, R0,R1,R2)
|
||||
IMPLICIT NONE
|
||||
double precision, dimension(:,:),intent(inout) :: BIG
|
||||
double precision, dimension(:),intent(in) :: R0,R1,R2
|
||||
integer ,intent(in) :: tn,ts
|
||||
integer :: i,j,Ntd1,N !=Ntdc
|
||||
double precision :: tmp
|
||||
! the order of the variables in the covariance matrix
|
||||
! are organized as follows:
|
||||
!
|
||||
! ||X(t2)..X(ts),..X(tn-1)||X'(ts) X'(t1) X'(tn)||Y1 Y2 X(ts) X(t1) X(tn)||
|
||||
! = [Xt Xd Xc]
|
||||
! where
|
||||
!
|
||||
! Xt= time points in the indicator function
|
||||
! Xd= derivatives
|
||||
! Xc=variables to condition on
|
||||
|
||||
! Computations of all covariances follows simple rules: Cov(X(t),X(s))=r(t,s),
|
||||
! then Cov(X'(t),X(s))=dr(t,s)/dt. Now for stationary X(t) we have
|
||||
! a function r(tau) such that Cov(X(t),X(s))=r(s-t) (or r(t-s) will give the same result).
|
||||
!
|
||||
! Consequently Cov(X'(t),X(s)) = -r'(s-t) = -sign(s-t)*r'(|s-t|)
|
||||
! Cov(X'(t),X'(s)) = -r''(s-t) = -r''(|s-t|)
|
||||
! Cov(X''(t),X'(s)) = r'''(s-t) = sign(s-t)*r'''(|s-t|)
|
||||
! Cov(X''(t),X(s)) = r''(s-t) = r''(|s-t|)
|
||||
! Cov(X''(t),X''(s)) = r''''(s-t) = r''''(|s-t|)
|
||||
|
||||
Ntd1=tn+1
|
||||
N=Ntd1+Nc
|
||||
do i=1,tn-2
|
||||
!cov(Xt)
|
||||
do j=i,tn-2
|
||||
BIG(i,j) = R0(j-i+1) ! cov(X(ti+1),X(tj+1))
|
||||
enddo
|
||||
!cov(Xt,Xc)
|
||||
BIG(i ,Ntd1+1) = 0.d0 !cov(X(ti+1),Y1)
|
||||
BIG(i ,Ntd1+2) = 0.d0 !cov(X(ti+1),Y2)
|
||||
BIG(i ,Ntd1+4) = R0(i+1) !cov(X(ti+1),X(t1))
|
||||
BIG(tn-1-i ,Ntd1+5) = R0(i+1) !cov(X(t.. ),X(tn))
|
||||
|
||||
!Cov(Xt,Xd)=cov(X(ti+1),x(tj)
|
||||
BIG(i,Ntd1-1) =-R1(i+1) !cov(X(ti+1),X'(t1))
|
||||
BIG(tn-1-i,Ntd1)= R1(i+1) !cov(X(ti+1),X'(tn))
|
||||
enddo
|
||||
!cov(Xd)
|
||||
BIG(Ntd1 ,Ntd1 ) = -R2(1)
|
||||
BIG(Ntd1-1,Ntd1 ) = -R2(tn) !cov(X'(t1),X'(tn))
|
||||
BIG(Ntd1-1,Ntd1-1) = -R2(1)
|
||||
BIG(Ntd1-2,Ntd1-1) = -R2(ts) !cov(X'(ts),X'(t1))
|
||||
BIG(Ntd1-2,Ntd1-2) = -R2(1)
|
||||
BIG(Ntd1-2,Ntd1 ) = -R2(tn+1-ts) !cov(X'(ts),X'(tn))
|
||||
|
||||
!cov(Xc)
|
||||
BIG(Ntd1+1,Ntd1+1) = 100.d0 ! cov(Y1 Y1)
|
||||
BIG(Ntd1+1,Ntd1+2) = 0.d0 ! cov(Y1 Y2)
|
||||
BIG(Ntd1+1,Ntd1+3) = 0.d0 ! cov(Y1 X(ts))
|
||||
BIG(Ntd1+1,Ntd1+4) = 0.d0 ! cov(Y1 X(t1))
|
||||
BIG(Ntd1+1,Ntd1+5) = 0.d0 ! cov(Y1 X(tn))
|
||||
BIG(Ntd1+2,Ntd1+2) = 100.d0 ! cov(Y2 Y2)
|
||||
BIG(Ntd1+2,Ntd1+3) = 0.d0 ! cov(Y2 X(ts))
|
||||
BIG(Ntd1+2,Ntd1+4) = 0.d0 ! cov(Y2 X(t1))
|
||||
BIG(Ntd1+2,Ntd1+5) = 0.d0 ! cov(Y2 X(tn))
|
||||
|
||||
BIG(Ntd1+3,Ntd1+3) = R0(1) ! cov(X(ts),X (ts)
|
||||
BIG(Ntd1+3,Ntd1+4) = R0(ts) ! cov(X(ts),X (t1))
|
||||
BIG(Ntd1+3,Ntd1+5) = R0(tn+1-ts) ! cov(X(ts),X (tn))
|
||||
BIG(Ntd1+4,Ntd1+4) = R0(1) ! cov(X(t1),X (t1))
|
||||
BIG(Ntd1+4,Ntd1+5) = R0(tn) ! cov(X(t1),X (tn))
|
||||
BIG(Ntd1+5,Ntd1+5) = R0(1) ! cov(X(tn),X (tn))
|
||||
|
||||
|
||||
!cov(Xd,Xc)
|
||||
BIG(Ntd1 ,Ntd1+1) = 0.d0 !cov(X'(tn),Y1)
|
||||
BIG(Ntd1 ,Ntd1+2) = 0.d0 !cov(X'(tn),Y2)
|
||||
BIG(Ntd1-1 ,Ntd1+1) = 0.d0 !cov(X'(t1),Y1)
|
||||
BIG(Ntd1-1 ,Ntd1+2) = 0.d0 !cov(X'(t1),Y2)
|
||||
BIG(Ntd1-2 ,Ntd1+1) = 0.d0 !cov(X'(ts),Y1)
|
||||
BIG(Ntd1-2 ,Ntd1+2) = 0.d0 !cov(X'(ts),Y2)
|
||||
|
||||
BIG(Ntd1 ,Ntd1+4) = R1(tn) !cov(X'(tn),X(t1))
|
||||
BIG(Ntd1 ,Ntd1+5) = 0.d0 !cov(X'(tn),X(tn))
|
||||
BIG(Ntd1-1,Ntd1+4) = 0.d0 !cov(X'(t1),X(t1))
|
||||
BIG(Ntd1-1,Ntd1+5) =-R1(tn) !cov(X'(t1),X(tn))
|
||||
BIG(Ntd1 ,Ntd1+3) = R1(tn+1-ts) !cov(X'(tn),X (ts))
|
||||
BIG(Ntd1-1,Ntd1+3) =-R1(ts) !cov(X'(t1),X (ts))
|
||||
BIG(Ntd1-2,Ntd1+3) = 0.d0 !cov(X'(ts),X (ts)
|
||||
BIG(Ntd1-2,Ntd1+4) = R1(ts) !cov(X'(ts),X (t1))
|
||||
BIG(Ntd1-2,Ntd1+5) = -R1(tn+1-ts) !cov(X'(ts),X (tn))
|
||||
|
||||
|
||||
do i=1,tn-2
|
||||
j=abs(i+1-ts)
|
||||
!cov(Xt,Xc)
|
||||
BIG(i,Ntd1+3) = R0(j+1) !cov(X(ti+1),X(ts))
|
||||
!Cov(Xt,Xd)
|
||||
if ((i+1-ts).lt.0) then
|
||||
BIG(i,Ntd1-2) = R1(j+1)
|
||||
else !cov(X(ti+1),X'(ts))
|
||||
BIG(i,Ntd1-2) = -R1(j+1)
|
||||
endif
|
||||
enddo
|
||||
|
||||
! make lower triangular part equal to upper
|
||||
do j=1,N-1
|
||||
do i=j+1,N
|
||||
tmp =BIG(j,i)
|
||||
|
||||
BIG(i,j)=tmp
|
||||
enddo
|
||||
enddo
|
||||
|
||||
C write (*,10) ((BIG(j,i),i=N+1,N+6),j=N+1,N+6)
|
||||
C 10 format(6F8.4)
|
||||
RETURN
|
||||
END SUBROUTINE COV_INPUT
|
||||
|
||||
SUBROUTINE COV_INPUT2(BIG,pt, R0,R1,R2)
|
||||
IMPLICIT NONE
|
||||
double precision, dimension(:,:), intent(out) :: BIG
|
||||
double precision, dimension(:), intent(in) :: R0,R1,R2
|
||||
integer :: pt,i,j
|
||||
! the order of the variables in the covariance matrix
|
||||
! are organized as follows;
|
||||
! X(t2)...X(tn-1) X'(t1) X'(tn) X(t1) X(tn) = [Xt Xd Xc]
|
||||
!
|
||||
! where Xd is the derivatives
|
||||
!
|
||||
! Xt= time points in the indicator function
|
||||
! Xd= derivatives
|
||||
! Xc=variables to condition on
|
||||
|
||||
!cov(Xc)
|
||||
BIG(pt+2,pt+2) = R0(1)
|
||||
BIG(pt+1,pt+1) = R0(1)
|
||||
BIG(pt+1,pt+2) = R0(pt)
|
||||
!cov(Xd)
|
||||
BIG(pt,pt) = -R2(1)
|
||||
BIG(pt-1,pt-1) = -R2(1)
|
||||
BIG(pt-1,pt) = -R2(pt)
|
||||
!cov(Xd,Xc)
|
||||
BIG(pt,pt+2) = 0.d0
|
||||
BIG(pt,pt+1) = R1(pt)
|
||||
BIG(pt-1,pt+2) = -R1(pt)
|
||||
BIG(pt-1,pt+1) = 0.d0
|
||||
|
||||
if (pt.GT.2) then
|
||||
!cov(Xt)
|
||||
do i=1,pt-2
|
||||
do j=i,pt-2
|
||||
BIG(i,j) = R0(j-i+1)
|
||||
enddo
|
||||
enddo
|
||||
!cov(Xt,Xc)
|
||||
do i=1,pt-2
|
||||
BIG(i,pt+1) = R0(i+1)
|
||||
BIG(pt-1-i,pt+2) = R0(i+1)
|
||||
enddo
|
||||
!Cov(Xt,Xd)=cov(X(ti+1),x(tj))
|
||||
do i=1,pt-2
|
||||
BIG(i,pt-1) = -R1(i+1)
|
||||
BIG(pt-1-i,pt)= R1(i+1)
|
||||
enddo
|
||||
endif
|
||||
|
||||
|
||||
! make lower triangular part equal to upper
|
||||
do j=1,pt+1
|
||||
do i=j+1,pt+2
|
||||
BIG(i,j)=BIG(j,i)
|
||||
enddo
|
||||
enddo
|
||||
C write (*,10) ((BIG(j,i),i=N+1,N+6),j=N+1,N+6)
|
||||
C 10 format(6F8.4)
|
||||
RETURN
|
||||
END SUBROUTINE COV_INPUT2
|
||||
|
||||
|
||||
END PROGRAM sp2tthpdf
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
@ -1,30 +0,0 @@
|
||||
"""
|
||||
f2py c_library.pyf c_functions.c -c
|
||||
|
||||
gfortran -W -Wall -pedantic-errors -fbounds-check -Werror -c dsvdc.f mregmodule.f
|
||||
|
||||
"""
|
||||
import os
|
||||
import sys
|
||||
from wafo.f2py_tools import f2py_call_str
|
||||
|
||||
|
||||
def compile_all():
|
||||
f2py_call = f2py_call_str()
|
||||
print '=' * 75
|
||||
print 'compiling cov2mod'
|
||||
print '=' * 75
|
||||
|
||||
files = ['dsvdc', 'mregmodule', 'intfcmod']
|
||||
compile1_format = 'gfortran -fPIC -c %s.f'
|
||||
format1 = '%s.o ' * len(files)
|
||||
for file_ in files:
|
||||
os.system(compile1_format % file_)
|
||||
file_objects = format1 % tuple(files)
|
||||
|
||||
os.system(f2py_call + ' -m cov2mod -c %s cov2mmpdfreg_intfc.f' %
|
||||
file_objects)
|
||||
|
||||
|
||||
if __name__ == '__main__':
|
||||
compile_all()
|
@ -1,651 +0,0 @@
|
||||
C Version 1994-X-18
|
||||
|
||||
C This is a new version of WAMP program computing crest-trough wavelength
|
||||
C and amplitude density.
|
||||
C
|
||||
C revised pab 2007
|
||||
C -moved all common blocks into modules
|
||||
C -renamed from minmax to sp2mmpdfreg + fixed some bugs
|
||||
C revised pab July 2007
|
||||
! -renamed from sp2mmpdfreg to cov2mmpdfreg
|
||||
|
||||
PROGRAM cov2mmpdfreg
|
||||
USE SIZEMOD
|
||||
USE EPSMOD
|
||||
USE CHECKMOD
|
||||
USE MREGMOD
|
||||
IMPLICIT NONE
|
||||
real*8 Q0,SQ0,Q1,SQ1, AA, BB, DAI, AI , U,V,VV, XL0, XL2, XL4
|
||||
REAL*8 VDERI, CDER,SDER, DER, CONST, F, HHHH,FM, VALUE
|
||||
C INTEGER, PARAMETER :: MMAX = 5, NMAX = 101, RDIM = 10201
|
||||
REAL*8, DIMENSION(NMAX) :: HHT,T,Ulev,Vlev,VT,UT,Vdd,Udd
|
||||
REAL*8, DIMENSION(RDIM) :: R,R1,R2,R3
|
||||
REAL*8, DIMENSION(5*NMAX) :: COV
|
||||
REAL*8, DIMENSION(NMAX,NMAX) :: UVdens
|
||||
C DIMENSION UVdens(NMAX,NMAX),HHT(NMAX)
|
||||
C DIMENSION T(NMAX),Ulev(NMAX),Vlev(NMAX)
|
||||
C DIMENSION VT(NMAX),UT(NMAX),Vdd(NMAX),Udd(NMAX)
|
||||
C DIMENSION COV(5*NMAX),R(RDIM),R1(RDIM),R2(RDIM),R3(RDIM)
|
||||
DIMENSION AA(MMAX-2,MMAX-2),BB(MMAX+1),DAI(MMAX),AI((MMAX+1)*NMAX)
|
||||
|
||||
C
|
||||
C The program computes the joint density of maximum the following minimum
|
||||
C and the distance between Max and min for a zero-mean stationary
|
||||
C Gaussian process with covariance function defined explicitely with 4
|
||||
C derivatives. The process should be normalized so that the first and
|
||||
C the second spectral moments are equal to 1. The values of Max are taken
|
||||
C as the nodes at Hermite-Quadrature and then integrated out so that
|
||||
C the output is a joint density of wavelength T and amplitude H=Max-min.
|
||||
C The Max values are defined by subroutine Gauss_M with the accuracy
|
||||
C input epsu. The principle is that the integral of the marginal density
|
||||
C of f_Max is computed with sufficient accuracy.
|
||||
C
|
||||
REAL*8, DIMENSION(NMAX) :: B0,DB0,DDB0,B1,DB1,DDB1,DB2,DDB2
|
||||
REAL*8, DIMENSION(NMAX) :: Q,SQ,VDER,DBI,BI
|
||||
C DIMENSION B0(NMAX),DB0(NMAX),DDB0(NMAX)
|
||||
C DIMENSION B1(NMAX),DB1(NMAX),DDB1(NMAX)
|
||||
C DIMENSION DB2(NMAX),DDB2(NMAX)
|
||||
C DIMENSION Q(NMAX),SQ(NMAX),VDER(NMAX),DBI(NMAX),BI(NMAX)
|
||||
INTEGER :: J,I,I1,I2,I3,IU, IV, NU,NV,NG,N,NIT, NNIT, INF
|
||||
INTEGER :: fffff
|
||||
C REAL*8 EPS0
|
||||
C INTEGER III01,III11,III21,III31,III41,III51
|
||||
C *,III61,III71,III81,III91,III101 , III0
|
||||
C COMMON/CHECK1/III01,III11,III21,III31,III41,III51
|
||||
C *,III61,III71,III81,III91,III101
|
||||
C COMMON/CHECKQ/III0
|
||||
C COMMON /EPS/ EPS,EPSS,CEPSS
|
||||
|
||||
C
|
||||
C Initiation of all constants and integration nodes 'INITINTEG'
|
||||
C
|
||||
CALL INITINTEG(NIT)
|
||||
c
|
||||
c OBS. we are using the variables R,R1,R2 R3 as a temporary storage
|
||||
C for transformation g of the process.
|
||||
|
||||
|
||||
|
||||
c
|
||||
CALL INITLEVELS(Ulev,NU,Vlev,NV,T,HHT,N,R1,R2,NG)
|
||||
IF( R1(1) .gt. R1(ng)) then
|
||||
do 13 I=1,ng
|
||||
R3(I)=R1(I)
|
||||
R(I) =R2(I)
|
||||
13 continue
|
||||
do 17 i=1,ng
|
||||
R1(i) = R3(ng-i+1)
|
||||
R2(i) = R(ng-i+1)
|
||||
17 continue
|
||||
end if
|
||||
if(abs(R1(ng)-R1(1))*abs(R2(ng)-R2(1)).lt.0.01d0) then
|
||||
print *,'The transformation g is singular, stop'
|
||||
stop
|
||||
end if
|
||||
DO 14 IV=1,Nv
|
||||
V=Vlev(IV)
|
||||
CALL TRANSF(NG,V,R2,R1,VALUE,DER)
|
||||
VT(IV)=VALUE
|
||||
Vdd(IV)=DER
|
||||
14 continue
|
||||
DO 16 IU=1,Nu
|
||||
U = Ulev(IU)
|
||||
CALL TRANSF(NG,U,R2,R1,VALUE,DER)
|
||||
UT(IU) = VALUE
|
||||
Udd(IU) = DER
|
||||
do 16 IV=1,Nv
|
||||
UVdens(IU,IV)=0.0d0
|
||||
16 CONTINUE
|
||||
|
||||
|
||||
CALL COVG(XL0,XL2,XL4,COV,R1,R2,R3,T,N)
|
||||
|
||||
|
||||
Q0=XL4
|
||||
IF (Q0.le.1.0D0+EPS) then
|
||||
Print *,'Covariance structure is singular, stop.'
|
||||
stop
|
||||
end if
|
||||
SQ0 = SQRT(Q0)
|
||||
Q1 = XL0-XL2*XL2/XL4
|
||||
IF (Q1.le.eps) then
|
||||
Print *,'Covariance structure is singular, stop.'
|
||||
stop
|
||||
end if
|
||||
SQ1 = SQRT(Q1)
|
||||
DO 10 I=1,N
|
||||
B0(I) =-COV(I+2*N)
|
||||
DB0(I) =-COV(I+3*N)
|
||||
DDB0(I)=-COV(I+4*N)
|
||||
|
||||
B1(I) =COV(I)+COV(I+2*N)*(XL2/XL4)
|
||||
DB1(I) =COV(I+N)+COV(I+3*N)*(XL2/XL4)
|
||||
DDB1(I)=COV(I+2*N)+XL2*(COV(I+4*N)/XL4)
|
||||
C
|
||||
C Q(I) contains Var(X(T(i))|X'(0),X''(0),X(0))
|
||||
C VDER(I) contains Var(X''(T(i))|X'(0),X''(0),X(0))
|
||||
C
|
||||
Q(I)=XL0 - COV(I+N)*(COV(I+N)/XL2) - B0(I)*(B0(I)/Q0)
|
||||
1 -B1(I)*(B1(I)/Q1)
|
||||
VDER(I)=XL4 - (COV(I+3*N)*COV(I+3*N))/XL2 - (DDB0(I)*DDB0(I))/Q0
|
||||
1 - (DDB1(I)*DDB1(I))/Q1
|
||||
|
||||
|
||||
C
|
||||
C DDB2(I) contains Cov(X''(T(i)),X(T(i))|X'(0),X''(0),X(0))
|
||||
C
|
||||
DDB2(I)=-XL2 - (COV(I+N)*COV(I+3*N))/XL2 - DDB0(I)*(B0(I)/Q0)
|
||||
1 -DDB1(I)*(B1(I)/Q1)
|
||||
IF(Q(I).LE.eps) then
|
||||
SQ(i) =0.0d0
|
||||
DDB2(i)=0.0d0
|
||||
else
|
||||
SQ(I)=SQRT(Q(I))
|
||||
C
|
||||
C VDER(I) contains Var(X''(T(i))|X'(0),X''(0),X(0),X(T(i))
|
||||
C
|
||||
|
||||
VDER(I)=VDER(I) - (DDB2(I)*DDB2(I))/Q(I)
|
||||
end if
|
||||
|
||||
10 CONTINUE
|
||||
DO 15 I=1,N
|
||||
DO 15 J=1,N
|
||||
C
|
||||
C R1 contains Cov(X(T(I)),X'(T(J))|X'(0),X''(0),X(0))
|
||||
C
|
||||
R1(J+(I-1)*N)=R1(J+(I-1)*N) - COV(I+N)*(COV(J+2*N)/XL2)
|
||||
1 - (B0(I)*DB0(J)/Q0) - (B1(I)*DB1(J)/Q1)
|
||||
|
||||
C
|
||||
C R2 contains Cov(X'(T(I)),X'(T(J))|X'(0),X''(0),X(0))
|
||||
C
|
||||
R2(J+(I-1)*N) = -R2(J+(I-1)*N) - COV(I+2*N)*(COV(J+2*N)/XL2)
|
||||
1 - DB0(I)*DB0(J)/Q0 - DB1(I)*(DB1(J)/Q1)
|
||||
C
|
||||
C R3 contains Cov(X''(T(I)),X'(T(J))|X'(0),X''(0),X(0))
|
||||
C
|
||||
R3(J+(I-1)*N) = R3(J+(I-1)*N) - COV(I+3*N)*(COV(J+2*N)/XL2)
|
||||
1 - DB0(J)*(DDB0(I)/Q0) - DDB1(I)*(DB1(J)/Q1)
|
||||
15 CONTINUE
|
||||
|
||||
C The initiations are finished and we are beginning with 3 loops
|
||||
C on T=T(I), U=Ulevels(IU), V=Ulevels(IV), U>V.
|
||||
|
||||
DO 20 I=1,N
|
||||
|
||||
NNIT=NIT
|
||||
IF (Q(I).LE.EPS) GO TO 20
|
||||
|
||||
DO 30 I1=1,I
|
||||
DB2(I1)=R1(I1+(I-1)*N)
|
||||
|
||||
C Cov(X'(T(I1)),X(T(i))|X'(0),X''(0),X(0))
|
||||
C DDB2(I) contains Cov(X''(T(i)),X(T(i))|X'(0),X''(0),X(0))
|
||||
|
||||
30 CONTINUE
|
||||
|
||||
DO 50 I3=1,I
|
||||
DBI(I3) = R3(I3+(I-1)*N) - (DDB2(I)*DB2(I3)/Q(I))
|
||||
BI(I3) = R2(I3+(I-1)*N) - (DB2(I)*DB2(I3)/Q(I))
|
||||
50 CONTINUE
|
||||
DO 51 I3=1,I-1
|
||||
AI(I3)=0.0d0
|
||||
AI(I3+I-1)=DB0(I3)/SQ0
|
||||
AI(I3+2*(I-1))=DB1(I3)/SQ1
|
||||
AI(I3+3*(I-1))=DB2(I3)/SQ(I)
|
||||
51 CONTINUE
|
||||
VDERI=VDER(I)
|
||||
DAI(1)=0.0d0
|
||||
DAI(2)=DDB0(I)/SQ0
|
||||
DAI(3)=DDB1(I)/SQ1
|
||||
DAI(4)=DDB2(I)/SQ(I)
|
||||
AA(1,1)=DB0(I)/SQ0
|
||||
AA(1,2)=DB1(I)/SQ1
|
||||
AA(1,3)=DB2(I)/SQ(I)
|
||||
AA(2,1)=XL2/SQ0
|
||||
AA(2,2)=SQ1
|
||||
AA(2,3)=0.0d0
|
||||
AA(3,1)=B0(I)/SQ0
|
||||
AA(3,2)=B1(I)/SQ1
|
||||
AA(3,3)=SQ(I)
|
||||
IF (BI(I).LE.EPS) NNIT=0
|
||||
IF (NNIT.GT.1) THEN
|
||||
IF(I.LT.1) GO TO 41
|
||||
DO 40 I1=1,I-1
|
||||
DO 40 I2=1,I-1
|
||||
|
||||
C R contains Cov(X'(T(I1)),X'(T(I2))|X'(0),X''(0),X(0),X(I))
|
||||
|
||||
R(I2+(I1-1)*(I-1))=R2(I2+(I1-1)*N)-(DB2(I1)*DB2(I2)/Q(I))
|
||||
|
||||
40 CONTINUE
|
||||
41 CONTINUE
|
||||
END IF
|
||||
|
||||
C Here the covariance of the problem would be innitiated
|
||||
|
||||
INF=0
|
||||
Print *,' Laps to go:',N-I+1
|
||||
DO 80 IV=1,Nv
|
||||
V=VT(IV)
|
||||
! IF (ABS(V).GT.5.0D0) GO TO 80
|
||||
IF (Vdd(IV).LT.EPS0) GO TO 80
|
||||
DO 60 IU=1,Nu
|
||||
U=UT(IU)
|
||||
IF (U.LE.V) go to 60
|
||||
! IF (ABS(U).GT.5.0D0) GO TO 60
|
||||
IF (Udd(IU).LT.EPS0) GO TO 60
|
||||
BB(1)=0.0d0
|
||||
BB(2)=U
|
||||
BB(3)=V
|
||||
! if (IV.EQ.2.AND.IU.EQ.1) THEN
|
||||
! fffff = 10
|
||||
! endif
|
||||
|
||||
CALL MREG(F,R,BI,DBI,AA,BB,AI,DAI,VDERI,3,I-1,NNIT,INF)
|
||||
INF=1
|
||||
UVdens(IU,IV) = UVdens(IU,IV) + Udd(IU)*Vdd(IV)*HHT(I)*F
|
||||
! if (F.GT.0.01.AND.U.GT.2.AND.V.LT.-2) THEN
|
||||
! if (N-I+1 .eq. 38.and.IV.EQ.26.AND.IU.EQ.16) THEN
|
||||
! if (IV.EQ.32.AND.IU.EQ.8.and.I.eq.11) THEN
|
||||
! PRINT * ,' R:', R(1:I)
|
||||
! PRINT * ,' BI:', BI(1:I)
|
||||
! PRINT * ,' DBI:', DBI(1:I)
|
||||
! PRINT * ,' DB2:', DB2(1:I)
|
||||
! PRINT * ,' DB0(1):', DB0(1)
|
||||
! PRINT * ,' DB1(1):', DB1(1)
|
||||
! PRINT * ,' DAI:', DAI
|
||||
! PRINT * ,' BB:', BB
|
||||
! PRINT * ,' VDERI:', VDERI
|
||||
! PRINT * ,' F :', F
|
||||
! PRINT * ,' UVDENS :', UVdens(IU,IV)
|
||||
! fffff = 10
|
||||
! endif
|
||||
|
||||
60 CONTINUE
|
||||
80 continue
|
||||
20 CONTINUE
|
||||
hhhh=0.0d0
|
||||
do 90 Iu=1,Nu
|
||||
do 90 Iv=1,Nv
|
||||
WRITE(10,300) Ulev(iu),Vlev(iv),UVdens(iu,iv)
|
||||
hhhh=hhhh+UVdens(iu,iv)
|
||||
90 continue
|
||||
if (nu.gt.1.and.nv.gt.1) then
|
||||
write(11,*) 'SumSum f_uv *du*dv='
|
||||
1,(Ulev(2)-Ulev(1))*(Vlev(2)-Vlev(1))*hhhh
|
||||
end if
|
||||
|
||||
C sder=sqrt(XL4-XL2*XL2/XL0)
|
||||
C cder=-XL2/sqrt(XL0)
|
||||
C const=1/sqrt(XL0*XL4)
|
||||
C DO 95 IU=1,NU
|
||||
C U=UT(IU)
|
||||
C FM=Udd(IU)*const*exp(-0.5*U*U/XL0)*PMEAN(-cder*U,sder)
|
||||
C WRITE(9,300) Ulev(IU),FM
|
||||
C 95 continue
|
||||
C DO 105 IV=1,NV
|
||||
C V=VT(IV)
|
||||
C VV=cder*V
|
||||
C Fm=Vdd(IV)*const*exp(-0.5*V*V/XL0)*PMEAN(VV,sder)
|
||||
C WRITE(8,300) Vlev(IV),Fm
|
||||
C 105 continue
|
||||
if (III0.eq.0) III0=1
|
||||
|
||||
write(11,*) 'Rate of calls RINDT0:',float(iii01)/float(III0)
|
||||
write(11,*) 'Rate of calls RINDT1:',float(iii11)/float(III0)
|
||||
write(11,*) 'Rate of calls RINDT2:',float(iii21)/float(III0)
|
||||
write(11,*) 'Rate of calls RINDT3:',float(iii31)/float(III0)
|
||||
write(11,*) 'Rate of calls RINDT4:',float(iii41)/float(III0)
|
||||
write(11,*) 'Rate of calls RINDT5:',float(iii51)/float(III0)
|
||||
write(11,*) 'Rate of calls RINDT6:',float(iii61)/float(III0)
|
||||
write(11,*) 'Rate of calls RINDT7:',float(iii71)/float(III0)
|
||||
write(11,*) 'Rate of calls RINDT8:',float(iii81)/float(III0)
|
||||
write(11,*) 'Rate of calls RINDT9:',float(iii91)/float(III0)
|
||||
write(11,*) 'Rate of calls RINDT10:',float(iii101)/float(III0)
|
||||
write(11,*) 'Number of calls of RINDT*',III0
|
||||
|
||||
CLOSE(UNIT=8)
|
||||
CLOSE(UNIT=9)
|
||||
CLOSE(UNIT=10)
|
||||
CLOSE(UNIT=11)
|
||||
|
||||
300 FORMAT(4(3X,F10.6))
|
||||
STOP
|
||||
END
|
||||
|
||||
SUBROUTINE INITLEVELS(ULEVELS,NU,Vlevels,Nv,T,HT,N,TG,XG,NG)
|
||||
USE TBRMOD
|
||||
USE SIZEMOD
|
||||
IMPLICIT NONE
|
||||
C INTEGER, PARAMETER:: NMAX = 101, RDIM = 10201
|
||||
C DIMENSION ULEVELS(1),Vlevels(1),T(1),HT(1),TG(1),XG(1),HH(101)
|
||||
REAL*8, DIMENSION(NMAX), intent(inout) :: ULEVELS,Vlevels,T,HT
|
||||
REAL*8, DIMENSION(RDIM), intent(inout) :: TG,XG
|
||||
INTEGER, intent(inout) :: NG
|
||||
REAL*8 :: UMIN,UMAX,VMIN,VMAX, HU,HV
|
||||
integer :: N, I, NU, NV
|
||||
C REAL*8, DIMENSION(NMAX) :: HH
|
||||
C COMMON/TBR/HH
|
||||
OPEN(UNIT=2,FILE='transf.in')
|
||||
OPEN(UNIT=4,FILE='Mm.in')
|
||||
OPEN(UNIT=3,FILE='t.in')
|
||||
|
||||
|
||||
NG=1
|
||||
12 READ (2,*,END=11) TG(NG),XG(NG)
|
||||
NG=NG+1
|
||||
GO TO 12
|
||||
11 CONTINUE
|
||||
NG=NG-1
|
||||
IF (NG.GT.501) THEN
|
||||
PRINT *,'Vector defining transformation of data > 501, stop'
|
||||
STOP
|
||||
END IF
|
||||
|
||||
|
||||
N=1
|
||||
32 READ (3,*,END=31) T(N)
|
||||
N=N+1
|
||||
GO TO 32
|
||||
31 CONTINUE
|
||||
N=N-1
|
||||
|
||||
CLOSE(UNIT=3)
|
||||
|
||||
IF(N.ge.NMAX) then
|
||||
print *,'The number of wavelength points >',NMAX-1, ' stop'
|
||||
stop
|
||||
end if
|
||||
IF(N.lt.2) then
|
||||
print *,'The number of wavelength points < 2, stop'
|
||||
stop
|
||||
end if
|
||||
|
||||
HT(1)=0.5d0*(T(2)-T(1))
|
||||
HT(N)=0.5d0*(T(N)-T(N-1))
|
||||
HH(1)=-100.0d0
|
||||
HH(N)=-100.0d0
|
||||
DO 10 I=2,N-1
|
||||
HT(I)=0.5d0*(T(I+1)-T(I-1))
|
||||
HH(I)=-100.0d0
|
||||
10 CONTINUE
|
||||
|
||||
|
||||
|
||||
READ(4,*) Umin,Umax,NU
|
||||
READ(4,*) Vmin,Vmax,NV
|
||||
|
||||
IF(NU.gt.NMAX) then
|
||||
print *,'The number of maxima >',NMAX,' stop'
|
||||
stop
|
||||
end if
|
||||
IF(NV.gt.NMAX) then
|
||||
print *,'The number of minima >',NMAX,' stop'
|
||||
stop
|
||||
end if
|
||||
|
||||
IF(NU.LT.1) Then
|
||||
print *,'The number of maxima < 1, stop'
|
||||
stop
|
||||
end if
|
||||
IF(NV.LT.1) Then
|
||||
print *,'The number of minima < 1, stop'
|
||||
stop
|
||||
end if
|
||||
|
||||
Ulevels(1)=Umax
|
||||
IF (NU.lt.2) go to 25
|
||||
HU=(Umax-Umin)/DBLE(NU-1)
|
||||
DO 20 I=1,NU-1
|
||||
ULEVELS(I+1)=Umax-DBLE(I)*HU
|
||||
20 CONTINUE
|
||||
|
||||
25 continue
|
||||
Vlevels(1)=Vmax
|
||||
IF (NV.lt.2) go to 35
|
||||
HV=(Vmax-Vmin)/DBLE(NV-1)
|
||||
DO 30 I=1,Nv-1
|
||||
VLEVELS(I+1)=Vmax-DBLE(I)*HV
|
||||
30 CONTINUE
|
||||
35 continue
|
||||
CLOSE(UNIT=4)
|
||||
RETURN
|
||||
END
|
||||
|
||||
|
||||
SUBROUTINE TRANSF(N,T,A,TIMEV,VALUE,DER)
|
||||
C
|
||||
C N number of data points
|
||||
C TIMEV vector of time points
|
||||
C A a vector of values of a function G(TIME)
|
||||
C T independent time point
|
||||
C VALUE is a value of a function at T, i.e. VALUE=G(T).
|
||||
c DER=G'(t)
|
||||
C
|
||||
USE SIZEMOD
|
||||
IMPLICIT NONE
|
||||
REAL*8, intent(inout):: VALUE, DER,T
|
||||
C INTEGER, PARAMETER :: RDIM = 10201
|
||||
REAL*8, DIMENSION(RDIM), intent(in) :: A,TIMEV
|
||||
integer, intent(in) :: N
|
||||
REAL*8:: T1
|
||||
integer :: I
|
||||
|
||||
IF (T.LT.TIMEV(1)) then
|
||||
der=(A(2)-A(1))/(TIMEV(2)-TIMEV(1))
|
||||
T1=T-TIMEV(1)
|
||||
VALUE=A(1)+T1*DER
|
||||
return
|
||||
end if
|
||||
IF (T.GT.TIMEV(N)) then
|
||||
der = (A(N)-A(N-1))/(TIMEV(N)-TIMEV(N-1))
|
||||
T1 = T-TIMEV(N)
|
||||
VALUE=A(N)+T1*DER
|
||||
return
|
||||
end if
|
||||
DO 5 I=2,N
|
||||
IF (T.LT.TIMEV(I)) GO TO 10
|
||||
5 CONTINUE
|
||||
10 I=I-1
|
||||
T1=T-TIMEV(I)
|
||||
DER=(A(I+1)-A(I))/(TIMEV(i+1)-TIMEV(I))
|
||||
VALUE=A(I)+T1*DER
|
||||
RETURN
|
||||
END
|
||||
|
||||
REAL*8 FUNCTION SPLE(N,T,A,TIMEV)
|
||||
C
|
||||
C N number of data points
|
||||
C TIME vector of time points
|
||||
C A a vector of values of a function G(TIME)
|
||||
C T independent time point
|
||||
C SPLE is a value of a function at T, i.e. SPLE=G(T).
|
||||
C
|
||||
USE SIZEMOD
|
||||
IMPLICIT NONE
|
||||
INTEGER, INTENT(IN):: N
|
||||
|
||||
REAL*8, INTENT(IN) :: T
|
||||
REAL*8, DIMENSION(5*NMAX), INTENT(IN) :: A,TIMEV
|
||||
REAL*8 :: T1
|
||||
INTEGER :: I
|
||||
SPLE=-9.9d0
|
||||
IF (T.LT.TIMEV(1) .OR. T.GT.TIMEV(N)) RETURN
|
||||
DO 5 I=2,N
|
||||
IF (T.LT.TIMEV(I)) GO TO 10
|
||||
5 CONTINUE
|
||||
10 I=I-1
|
||||
T1=T-TIMEV(I)
|
||||
SPLE=A(I)+T1*(A(I+1)-A(I))/(TIMEV(i+1)-TIMEV(I))
|
||||
RETURN
|
||||
END
|
||||
|
||||
|
||||
|
||||
SUBROUTINE COVG(XL0,XL2,XL4,COV,COV1,COV2,COV3,T,N)
|
||||
C
|
||||
C COVG evaluates:
|
||||
C
|
||||
C XL0,XL2,XL4 - spectral moments.
|
||||
C
|
||||
C Covariance function and its four derivatives for a vector T of length N.
|
||||
C It is saved in a vector COV; COV(1,...,N)=r(T), COV(N+1,...,2N)=r'(T), etc.
|
||||
C The vector COV should be of the length 5*N.
|
||||
C
|
||||
C Covariance matrices COV1=r'(T-T), COV2=r''(T-T) and COV3=r'''(T-T)
|
||||
C Dimension of COV1, COV2 should be N*N.
|
||||
C
|
||||
USE SIZEMOD
|
||||
! IMPLICIT NONE
|
||||
C INTEGER, PARAMETER:: NMAX = 101, RDIM = 10201
|
||||
REAL*8, PARAMETER:: ZERO = 0.0d0
|
||||
REAL*8, intent(inout) :: XL0,XL2,XL4
|
||||
REAL*8, DIMENSION(5*NMAX), intent(inout) :: COV
|
||||
REAL*8, DIMENSION(5*NMAX) :: A, TIMEV
|
||||
REAL*8, DIMENSION(RDIM), intent(inout) :: COV1,COV2,COV3
|
||||
REAL*8, DIMENSION(NMAX), intent(in) :: T
|
||||
INTEGER, intent(in) :: N
|
||||
integer :: NT, I, J, II
|
||||
REAL*8 :: TT, T0
|
||||
OPEN(UNIT=32,FILE='Cd0.in')
|
||||
OPEN(UNIT=33,FILE='Cd1.in')
|
||||
OPEN(UNIT=34,FILE='Cd2.in')
|
||||
OPEN(UNIT=35,FILE='Cd3.in')
|
||||
OPEN(UNIT=36,FILE='Cd4.in')
|
||||
C
|
||||
C COV(Y(T),Y(0))
|
||||
C
|
||||
|
||||
NT=1
|
||||
12 READ (32,*,END=11) TIMEV(NT),A(NT)
|
||||
NT=NT+1
|
||||
GO TO 12
|
||||
11 CONTINUE
|
||||
NT=NT-1
|
||||
|
||||
|
||||
XL0=SPLE(NT,ZERO,A,TIMEV)
|
||||
|
||||
DO 10 I=1,N
|
||||
COV(I)=SPLE(NT,T(I),A,TIMEV)
|
||||
10 CONTINUE
|
||||
|
||||
C
|
||||
C DERIVATIVE COV(Y(T),Y(0))
|
||||
C
|
||||
|
||||
NT=1
|
||||
22 READ (33,*,END=21) TIMEV(NT),A(NT)
|
||||
NT=NT+1
|
||||
GO TO 22
|
||||
21 CONTINUE
|
||||
NT=NT-1
|
||||
|
||||
II=0
|
||||
DO 20 I=1,N
|
||||
COV(I+N)=SPLE(NT,T(I),A,TIMEV)
|
||||
DO 20 J=1,N
|
||||
II=II+1
|
||||
T0=T(J)-T(I)
|
||||
TT=ABS(T0)
|
||||
COV1(II)=SPLE(NT,TT,A,TIMEV)
|
||||
IF (T0.LT.0.0d0) COV1(II)=-COV1(II)
|
||||
20 CONTINUE
|
||||
|
||||
C 2-DERIVATIVE COV(Y(T),Y(0))
|
||||
|
||||
NT=1
|
||||
32 READ (34,*,END=31) TIMEV(NT),A(NT)
|
||||
NT=NT+1
|
||||
GO TO 32
|
||||
31 CONTINUE
|
||||
NT=NT-1
|
||||
|
||||
II=0
|
||||
XL2=-SPLE(NT,ZERO,A,TIMEV)
|
||||
|
||||
DO 30 I=1,N
|
||||
COV(I+2*N)=SPLE(NT,T(I),A,TIMEV)
|
||||
DO 30 J=1,N
|
||||
II=II+1
|
||||
TT=ABS(T(J)-T(I))
|
||||
COV2(II)=SPLE(NT,TT,A,TIMEV)
|
||||
30 CONTINUE
|
||||
|
||||
C 3-DERIVATIVE COV(Y(T),Y(0))
|
||||
|
||||
NT=1
|
||||
42 READ (35,*,END=41) TIMEV(NT),A(NT)
|
||||
NT=NT+1
|
||||
GO TO 42
|
||||
41 CONTINUE
|
||||
NT=NT-1
|
||||
|
||||
|
||||
II=0
|
||||
DO 40 I=1,N
|
||||
COV(I+3*N)=SPLE(NT,T(I),A,TIMEV)
|
||||
DO 40 J=1,N
|
||||
II=II+1
|
||||
T0=T(J)-T(I)
|
||||
TT=ABS(T0)
|
||||
COV3(II)=SPLE(NT,TT,A,TIMEV)
|
||||
IF (T0.LT.0.0d0) COV3(II)=-COV3(II)
|
||||
40 CONTINUE
|
||||
|
||||
|
||||
|
||||
C 4-DERIVATIVE COV(Y(T),Y(0))
|
||||
|
||||
NT=1
|
||||
52 READ (36,*,END=51) TIMEV(NT),A(NT)
|
||||
NT=NT+1
|
||||
GO TO 52
|
||||
51 CONTINUE
|
||||
NT=NT-1
|
||||
|
||||
XL4=SPLE(NT,ZERO,A,TIMEV)
|
||||
|
||||
DO 50 I=1,N
|
||||
COV(I+4*N)=SPLE(NT,T(I),A,TIMEV)
|
||||
50 CONTINUE
|
||||
CLOSE(UNIT=32)
|
||||
CLOSE(UNIT=33)
|
||||
CLOSE(UNIT=34)
|
||||
CLOSE(UNIT=35)
|
||||
CLOSE(UNIT=36)
|
||||
RETURN
|
||||
END
|
||||
|
||||
SUBROUTINE INITINTEG(NIT)
|
||||
USE RINTMOD
|
||||
USE EPSMOD
|
||||
USE INFCMOD
|
||||
USE MREGMOD
|
||||
! IMPLICIT NONE
|
||||
INTEGER, intent(inout) :: NIT
|
||||
! INTEGER ISQ1
|
||||
C dimension INF(10),INFO(10)
|
||||
|
||||
C COMMON /RINT/ C,FC
|
||||
C COMMON /EPS/ EPS,EPSS,CEPSS
|
||||
C COMMON /INFC/ ISQ,INF,INFO
|
||||
OPEN(UNIT=1,FILE='accur.in')
|
||||
OPEN(UNIT=8,FILE='min.out')
|
||||
OPEN(UNIT=9,FILE='Max.out')
|
||||
OPEN(UNIT=10,FILE='Maxmin.out')
|
||||
OPEN(UNIT=11,FILE='Maxmin.log')
|
||||
|
||||
READ(1,*) NIT,IAC,ISQ
|
||||
READ(1,*) EPS,EPSS,EPS0
|
||||
|
||||
CLOSE (UNIT=1)
|
||||
|
||||
FC=FI(C)-FI(-C)
|
||||
CEPSS=1.0d0-EPSS
|
||||
|
||||
RETURN
|
||||
END
|
||||
|
@ -1,370 +0,0 @@
|
||||
C Version 1994-X-18
|
||||
|
||||
C This is a new version of WAMP program computing crest-trough wavelength
|
||||
C and amplitude density.
|
||||
C
|
||||
C revised pab 2007
|
||||
C -moved all common blocks into modules
|
||||
C -renamed from minmax to sp2mmpdfreg + fixed some bugs
|
||||
C revised pab July 2007
|
||||
! -renamed from sp2mmpdfreg to cov2mmpdfreg
|
||||
! gfortran -W -Wall -pedantic-errors -fbounds-check -Werror -c dsvdc.f mregmodule.f cov2mmpdfreg.f
|
||||
|
||||
SUBROUTINE INITINTEG(EPS_,EPSS_,EPS0_,C_,IAC_,ISQ_)
|
||||
! Initiation of all constants and integration nodes 'INITINTEG'
|
||||
USE RINTMOD
|
||||
USE EPSMOD
|
||||
USE INFCMOD
|
||||
USE MREGMOD
|
||||
REAL*8 :: EPS_,EPSS_,EPS0_,C_
|
||||
INTEGER :: IAC_,ISQ_
|
||||
Cf2py real*8, optional :: EPS_ = 0.01
|
||||
Cf2py real*8, optional :: EPSS_ = 0.00005
|
||||
Cf2py real*8, optional :: EPS0_ = 0.00005
|
||||
Cf2py real*8, optional :: C_ = 4.5
|
||||
Cf2py integer, optional :: IAC_ = 1
|
||||
Cf2py integer, optional :: ISQ_ = 0
|
||||
! IMPLICIT NONE
|
||||
C COMMON /RINT/ C,FC
|
||||
C COMMON /EPS/ EPS,EPSS,CEPSS
|
||||
C COMMON /INFC/ ISQ,INF,INFO
|
||||
|
||||
IAC = IAC_
|
||||
ISQ = ISQ_
|
||||
EPS = EPS_
|
||||
EPSS = EPSS_
|
||||
EPS0 = EPS0_
|
||||
C = C_
|
||||
|
||||
FC = FI(C)-FI(-C)
|
||||
! CEPSS = 1.0d0-EPSS
|
||||
RETURN
|
||||
END SUBROUTINE INITINTEG
|
||||
|
||||
subroutine cov2mmpdfreg(UVdens,t,COV,ULev,VLev,Tg,Xg,Nt,Nu,Nv,Ng,
|
||||
& NIT)
|
||||
USE SIZEMOD
|
||||
USE EPSMOD
|
||||
USE CHECKMOD
|
||||
USE MREGMOD
|
||||
USE INTFCMOD
|
||||
IMPLICIT NONE
|
||||
INTEGER, INTENT(IN) :: Nt, Nu, Nv, Ng, NIT
|
||||
REAL*8, DIMENSION(Nt,5), intent(in):: COV
|
||||
REAL*8, DIMENSION(Nu,Nv), intent(out):: UVdens
|
||||
REAL*8, DIMENSION(Nu), intent(in):: ULev
|
||||
REAL*8, DIMENSION(Nv), intent(in):: VLev
|
||||
REAL*8, DIMENSION(Ng), intent(in):: Tg, Xg
|
||||
REAL*8, dimension(Nt), intent(in):: T
|
||||
Cf2py integer, intent(hide), depend(t) :: Nt = len(t)
|
||||
Cf2py integer, intent(hide), depend(Ulev) :: Nu = len(Ulev)
|
||||
Cf2py integer, intent(hide), depend(Vlev) :: Nv = len(Vlev)
|
||||
Cf2py integer, intent(hide), depend(Tg) :: Ng = len(Tg)
|
||||
Cf2py integer, optional :: NIT = 2
|
||||
Cf2py real*8, intent(out), depend(Nu,Nv) :: UVdens
|
||||
Cf2py depend(Ng) Xg
|
||||
Cf2py depend(Nt,5) COV
|
||||
real*8 Q0,SQ0,Q1,SQ1, U,V, XL0, XL2, XL4
|
||||
REAL*8 VDERI, DER, F, HHHH, VALUE
|
||||
C REAL*8 VV, CDER,SDER, CONST1, FM
|
||||
C INTEGER, PARAMETER :: MMAX = 5, NMAX = 101, RDIM = 10201
|
||||
REAL*8, DIMENSION(NMAX) :: HHT,VT,UT,Vdd,Udd
|
||||
REAL*8, DIMENSION(RDIM) :: R,R1,R2,R3
|
||||
REAL*8:: AA(MMAX-2,MMAX-2),AI((MMAX+1)*NMAX)
|
||||
REAL*8, DIMENSION(MMAX+1) :: BB, DAI
|
||||
C DIMENSION UVdens(NMAX,NMAX),HHT(NMAX)
|
||||
C DIMENSION T(NMAX),Ulev(NMAX),Vlev(NMAX)
|
||||
C DIMENSION VT(NMAX),UT(NMAX),Vdd(NMAX),Udd(NMAX)
|
||||
C DIMENSION COV(5*NMAX),R(RDIM),R1(RDIM),R2(RDIM),R3(RDIM)
|
||||
|
||||
|
||||
C
|
||||
C The program computes the joint density of maximum the following minimum
|
||||
C and the distance between Max and min for a zero-mean stationary
|
||||
C Gaussian process with covariance function defined explicitely with 4
|
||||
C derivatives. The process should be normalized so that the first and
|
||||
C the second spectral moments are equal to 1. The values of Max are taken
|
||||
C as the nodes at Hermite-Quadrature and then integrated out so that
|
||||
C the output is a joint density of wavelength T and amplitude H=Max-min.
|
||||
C The Max values are defined by subroutine Gauss_M with the accuracy
|
||||
C input epsu. The principle is that the integral of the marginal density
|
||||
C of f_Max is computed with sufficient accuracy.
|
||||
C
|
||||
REAL*8, DIMENSION(NMAX) :: B0,DB0,DDB0,B1,DB1,DDB1,DB2,DDB2
|
||||
REAL*8, DIMENSION(NMAX) :: Q,SQ,VDER,DBI,BI
|
||||
C DIMENSION B0(NMAX),DB0(NMAX),DDB0(NMAX)
|
||||
C DIMENSION B1(NMAX),DB1(NMAX),DDB1(NMAX)
|
||||
C DIMENSION DB2(NMAX),DDB2(NMAX)
|
||||
C DIMENSION Q(NMAX),SQ(NMAX),VDER(NMAX),DBI(NMAX),BI(NMAX)
|
||||
INTEGER :: J,I,I1,I2,I3,IU, IV,N, NNIT, INF
|
||||
C INTEGER :: fffff
|
||||
C REAL*8 EPS0
|
||||
C INTEGER III01,III11,III21,III31,III41,III51
|
||||
C *,III61,III71,III81,III91,III101 , III0
|
||||
C COMMON/CHECK1/III01,III11,III21,III31,III41,III51
|
||||
C *,III61,III71,III81,III91,III101
|
||||
C COMMON/CHECKQ/III0
|
||||
C COMMON /EPS/ EPS,EPSS,CEPSS
|
||||
|
||||
C
|
||||
C Initiation of all constants and integration nodes 'INITINTEG'
|
||||
C
|
||||
! CALL INITINTEG()
|
||||
|
||||
! OPEN(UNIT=8,FILE='min.out')
|
||||
! OPEN(UNIT=9,FILE='Max.out')
|
||||
! OPEN(UNIT=10,FILE='Maxmin.out')
|
||||
! OPEN(UNIT=11,FILE='Maxmin.log')
|
||||
c
|
||||
c OBS. we are using the variables R,R1,R2 R3 as a temporary storage
|
||||
C for transformation g of the process.
|
||||
|
||||
N = Nt
|
||||
CALL INITLEVELS(T,HHT,Nt,NU,Nv)
|
||||
C CALL INITLEVELS(Ulev,NU,Vlev,NV,T,HHT,Nt,R1,R2,NG)
|
||||
IF( Tg(1) .gt. Tg(ng)) then
|
||||
print *,'Error Tg must be strictly increasing'
|
||||
return
|
||||
end if
|
||||
if(abs(Tg(ng)-Tg(1))*abs(Xg(ng)-Xg(1)).lt.0.01d0) then
|
||||
print *,'The transformation g is singular, stop'
|
||||
return
|
||||
end if
|
||||
|
||||
! do IV=1,Nt
|
||||
! print *, 'Cov', COV(IV,:)
|
||||
! end do
|
||||
|
||||
DO IV=1,Nv
|
||||
V=Vlev(IV)
|
||||
CALL TRANSF(NG,V,Xg,Tg,VALUE,DER)
|
||||
VT(IV)=VALUE
|
||||
Vdd(IV)=DER
|
||||
enddo
|
||||
DO IU=1,Nu
|
||||
U = Ulev(IU)
|
||||
CALL TRANSF(NG,U,Xg,Tg,VALUE,DER)
|
||||
UT(IU) = VALUE
|
||||
Udd(IU) = DER
|
||||
do IV=1,Nv
|
||||
UVdens(IU,IV)=0.0d0
|
||||
enddo
|
||||
enddo
|
||||
|
||||
CALL COVG(XL0,XL2,XL4,R1,R2,R3,COV,T,Nt)
|
||||
|
||||
|
||||
Q0=XL4
|
||||
IF (Q0.le.1.0D0+EPS) then
|
||||
Print *,'Covariance structure is singular, stop.'
|
||||
return
|
||||
end if
|
||||
SQ0 = SQRT(Q0)
|
||||
Q1 = XL0-XL2*XL2/XL4
|
||||
IF (Q1.le.EPS) then
|
||||
Print *,'Covariance structure is singular, stop.'
|
||||
return
|
||||
end if
|
||||
SQ1 = SQRT(Q1)
|
||||
DO I=1,Nt
|
||||
B0(I) =-COV(I,3)
|
||||
DB0(I) =-COV(I,4)
|
||||
DDB0(I)=-COV(I,5)
|
||||
|
||||
B1(I) =COV(I,1)+COV(I,3)*(XL2/XL4)
|
||||
DB1(I) =COV(I,2)+COV(I,4)*(XL2/XL4)
|
||||
DDB1(I)=COV(I,3)+XL2*(COV(I,5)/XL4)
|
||||
C
|
||||
C Q(I) contains Var(X(T(i))|X'(0),X''(0),X(0))
|
||||
C VDER(I) contains Var(X''(T(i))|X'(0),X''(0),X(0))
|
||||
C
|
||||
Q(I)=XL0 - COV(I,2)*(COV(I,2)/XL2) - B0(I)*(B0(I)/Q0)
|
||||
1 -B1(I)*(B1(I)/Q1)
|
||||
VDER(I)=XL4 - (COV(I,4)*COV(I,4))/XL2 - (DDB0(I)*DDB0(I))/Q0
|
||||
1 - (DDB1(I)*DDB1(I))/Q1
|
||||
|
||||
|
||||
C
|
||||
C DDB2(I) contains Cov(X''(T(i)),X(T(i))|X'(0),X''(0),X(0))
|
||||
C
|
||||
DDB2(I)=-XL2 - (COV(I,2)*COV(I,4))/XL2 - DDB0(I)*(B0(I)/Q0)
|
||||
1 -DDB1(I)*(B1(I)/Q1)
|
||||
IF(Q(I).LE.eps) then
|
||||
SQ(i) =0.0d0
|
||||
DDB2(i)=0.0d0
|
||||
else
|
||||
SQ(I)=SQRT(Q(I))
|
||||
C
|
||||
C VDER(I) contains Var(X''(T(i))|X'(0),X''(0),X(0),X(T(i))
|
||||
C
|
||||
|
||||
VDER(I)=VDER(I) - (DDB2(I)*DDB2(I))/Q(I)
|
||||
end if
|
||||
|
||||
c10 CONTINUE
|
||||
enddo
|
||||
DO I=1,Nt
|
||||
DO J=1,Nt
|
||||
C
|
||||
C R1 contains Cov(X(T(I)),X'(T(J))|X'(0),X''(0),X(0))
|
||||
C
|
||||
R1(J+(I-1)*N) = R1(J+(I-1)*N) - COV(I,2)*(COV(J,3)/XL2)
|
||||
1 - (B0(I)*DB0(J)/Q0) - (B1(I)*DB1(J)/Q1)
|
||||
|
||||
C
|
||||
C R2 contains Cov(X'(T(I)),X'(T(J))|X'(0),X''(0),X(0))
|
||||
C
|
||||
R2(J+(I-1)*N) = -R2(J+(I-1)*N) - COV(I,3)*(COV(J,3)/XL2)
|
||||
1 - DB0(I)*DB0(J)/Q0 - DB1(I)*(DB1(J)/Q1)
|
||||
C
|
||||
C R3 contains Cov(X''(T(I)),X'(T(J))|X'(0),X''(0),X(0))
|
||||
C
|
||||
R3(J+(I-1)*N) = R3(J+(I-1)*N) - COV(I,4)*(COV(J,3)/XL2)
|
||||
1 - DB0(J)*(DDB0(I)/Q0) - DDB1(I)*(DB1(J)/Q1)
|
||||
c15 CONTINUE
|
||||
enddo
|
||||
enddo
|
||||
|
||||
C The initiations are finished and we are beginning with 3 loops
|
||||
C on T=T(I), U=Ulevels(IU), V=Ulevels(IV), U>V.
|
||||
|
||||
DO I=1,Nt
|
||||
|
||||
NNIT=NIT
|
||||
IF (Q(I).LE.EPS) GO TO 20
|
||||
|
||||
DO I1=1,I
|
||||
DB2(I1)=R1(I1+(I-1)*N)
|
||||
|
||||
C Cov(X'(T(I1)),X(T(i))|X'(0),X''(0),X(0))
|
||||
C DDB2(I) contains Cov(X''(T(i)),X(T(i))|X'(0),X''(0),X(0))
|
||||
|
||||
enddo
|
||||
|
||||
DO I3=1,I
|
||||
DBI(I3) = R3(I3+(I-1)*N) - (DDB2(I)*DB2(I3)/Q(I))
|
||||
BI(I3) = R2(I3+(I-1)*N) - (DB2(I)*DB2(I3)/Q(I))
|
||||
enddo
|
||||
DO I3=1,I-1
|
||||
AI(I3)=0.0d0
|
||||
AI(I3+I-1)=DB0(I3)/SQ0
|
||||
AI(I3+2*(I-1))=DB1(I3)/SQ1
|
||||
AI(I3+3*(I-1))=DB2(I3)/SQ(I)
|
||||
enddo
|
||||
VDERI=VDER(I)
|
||||
DAI(1)=0.0d0
|
||||
DAI(2)=DDB0(I)/SQ0
|
||||
DAI(3)=DDB1(I)/SQ1
|
||||
DAI(4)=DDB2(I)/SQ(I)
|
||||
AA(1,1)=DB0(I)/SQ0
|
||||
AA(1,2)=DB1(I)/SQ1
|
||||
AA(1,3)=DB2(I)/SQ(I)
|
||||
AA(2,1)=XL2/SQ0
|
||||
AA(2,2)=SQ1
|
||||
AA(2,3)=0.0d0
|
||||
AA(3,1)=B0(I)/SQ0
|
||||
AA(3,2)=B1(I)/SQ1
|
||||
AA(3,3)=SQ(I)
|
||||
IF (BI(I).LE.EPS) NNIT=0
|
||||
IF (NNIT.GT.1) THEN
|
||||
IF(I.LT.1) GO TO 41
|
||||
DO I1=1,I-1
|
||||
DO I2=1,I-1
|
||||
C R contains Cov(X'(T(I1)),X'(T(I2))|X'(0),X''(0),X(0),X(I))
|
||||
R(I2+(I1-1)*(I-1))=R2(I2+(I1-1)*N)-(DB2(I1)*DB2(I2)/Q(I))
|
||||
|
||||
enddo
|
||||
enddo
|
||||
41 CONTINUE
|
||||
END IF
|
||||
|
||||
C Here the covariance of the problem would be initiated
|
||||
|
||||
INF=0
|
||||
Print *,' Laps to go:',Nt-I+1
|
||||
DO IV=1,Nv
|
||||
V=VT(IV)
|
||||
! IF (ABS(V).GT.5.0D0) GO TO 80
|
||||
IF (Vdd(IV).LT.EPS0) GO TO 80
|
||||
DO IU=1,Nu
|
||||
U=UT(IU)
|
||||
IF (U.LE.V) go to 60
|
||||
! IF (ABS(U).GT.5.0D0) GO TO 60
|
||||
IF (Udd(IU).LT.EPS0) GO TO 60
|
||||
BB(1)=0.0d0
|
||||
BB(2)=U
|
||||
BB(3)=V
|
||||
! if (IV.EQ.2.AND.IU.EQ.1) THEN
|
||||
! fffff = 10
|
||||
! endif
|
||||
|
||||
CALL MREG(F,R,BI,DBI,AA,BB,AI,DAI,VDERI,3,I-1,NNIT,INF)
|
||||
INF=1
|
||||
UVdens(IU,IV) = UVdens(IU,IV) + Udd(IU)*Vdd(IV)*HHT(I)*F
|
||||
|
||||
! if (F.GT.0.01.AND.U.GT.2.AND.V.LT.-2) THEN
|
||||
! if (N-I+1 .eq. 38.and.IV.EQ.26.AND.IU.EQ.16) THEN
|
||||
! if (IV.EQ.32.AND.IU.EQ.8.and.I.eq.11) THEN
|
||||
! PRINT * ,' R:', R(1:I)
|
||||
! PRINT * ,' BI:', BI(1:I)
|
||||
! PRINT * ,' DBI:', DBI(1:I)
|
||||
! PRINT * ,' DB2:', DB2(1:I)
|
||||
! PRINT * ,' DB0(1):', DB0(1)
|
||||
! PRINT * ,' DB1(1):', DB1(1)
|
||||
! PRINT * ,' DAI:', DAI
|
||||
! PRINT * ,' BB:', BB
|
||||
! PRINT * ,' VDERI:', VDERI
|
||||
! PRINT * ,' F :', F
|
||||
! PRINT * ,' UVDENS :', UVdens(IU,IV)
|
||||
! fffff = 10
|
||||
! endif
|
||||
|
||||
60 CONTINUE
|
||||
enddo
|
||||
80 continue
|
||||
enddo
|
||||
20 CONTINUE
|
||||
enddo
|
||||
|
||||
hhhh=0.0d0
|
||||
do Iu=1,Nu
|
||||
do Iv=1,Nv
|
||||
! WRITE(10,300) Ulev(iu),Vlev(iv),UVdens(iu,iv)
|
||||
hhhh=hhhh+UVdens(iu,iv)
|
||||
enddo
|
||||
enddo
|
||||
if (nu.gt.1.and.nv.gt.1) then
|
||||
VALUE = (Ulev(2)-Ulev(1))*(Vlev(2)-Vlev(1))*hhhh
|
||||
print *,'SumSum f_uv *du*dv=', VALUE
|
||||
end if
|
||||
|
||||
C sder=sqrt(XL4-XL2*XL2/XL0)
|
||||
C cder=-XL2/sqrt(XL0)
|
||||
C const1=1/sqrt(XL0*XL4)
|
||||
C DO 95 IU=1,NU
|
||||
C U=UT(IU)
|
||||
C FM=Udd(IU)*const1*exp(-0.5*U*U/XL0)*PMEAN(-cder*U,sder)
|
||||
C WRITE(9,300) Ulev(IU),FM
|
||||
C 95 continue
|
||||
C DO 105 IV=1,NV
|
||||
C V=VT(IV)
|
||||
C VV=cder*V
|
||||
C Fm=Vdd(IV)*const1*exp(-0.5*V*V/XL0)*PMEAN(VV,sder)
|
||||
C WRITE(8,300) Vlev(IV),Fm
|
||||
C 105 continue
|
||||
if (III0.eq.0) III0=1
|
||||
|
||||
PRINT *, 'Rate of calls RINDT0:',float(iii01)/float(III0)
|
||||
PRINT *, 'Rate of calls RINDT1:',float(iii11)/float(III0)
|
||||
PRINT *, 'Rate of calls RINDT2:',float(iii21)/float(III0)
|
||||
PRINT *, 'Rate of calls RINDT3:',float(iii31)/float(III0)
|
||||
PRINT *, 'Rate of calls RINDT4:',float(iii41)/float(III0)
|
||||
PRINT *, 'Rate of calls RINDT5:',float(iii51)/float(III0)
|
||||
PRINT *, 'Rate of calls RINDT6:',float(iii61)/float(III0)
|
||||
PRINT *, 'Rate of calls RINDT7:',float(iii71)/float(III0)
|
||||
PRINT *, 'Rate of calls RINDT8:',float(iii81)/float(III0)
|
||||
PRINT *, 'Rate of calls RINDT9:',float(iii91)/float(III0)
|
||||
PRINT *, 'Rate of calls RINDT10:',float(iii101)/float(III0)
|
||||
PRINT *, 'Number of calls of RINDT*',III0
|
||||
return
|
||||
END subroutine cov2mmpdfreg
|
@ -1,613 +0,0 @@
|
||||
MODULE SVD
|
||||
IMPLICIT NONE
|
||||
INTEGER, PARAMETER :: dp = SELECTED_REAL_KIND(12, 60)
|
||||
|
||||
! Based upon routines from the NSWC (Naval Surface Warfare Center),
|
||||
! which were based upon LAPACK routines.
|
||||
|
||||
! Code converted using TO_F90 by Alan Miller
|
||||
! Date: 2003-11-11 Time: 17:50:44
|
||||
! Revised pab 2007
|
||||
! Converted to fixed form
|
||||
|
||||
|
||||
CONTAINS
|
||||
|
||||
|
||||
SUBROUTINE drotg(da, db, dc, ds)
|
||||
|
||||
! DESIGNED BY C.L.LAWSON, JPL, 1977 SEPT 08
|
||||
!
|
||||
! CONSTRUCT THE GIVENS TRANSFORMATION
|
||||
!
|
||||
! ( DC DS )
|
||||
! G = ( ) , DC**2 + DS**2 = 1 ,
|
||||
! (-DS DC )
|
||||
!
|
||||
! WHICH ZEROS THE SECOND ENTRY OF THE 2-VECTOR (DA,DB)**T .
|
||||
!
|
||||
! THE QUANTITY R = (+/-)SQRT(DA**2 + DB**2) OVERWRITES DA IN
|
||||
! STORAGE. THE VALUE OF DB IS OVERWRITTEN BY A VALUE Z WHICH
|
||||
! ALLOWS DC AND DS TO BE RECOVERED BY THE FOLLOWING ALGORITHM:
|
||||
! IF Z=1 SET DC=0.D0 AND DS=1.D0
|
||||
! IF DABS(Z) < 1 SET DC=SQRT(1-Z**2) AND DS=Z
|
||||
! IF DABS(Z) > 1 SET DC=1/Z AND DS=SQRT(1-DC**2)
|
||||
!
|
||||
! NORMALLY, THE SUBPROGRAM DROT(N,DX,INCX,DY,INCY,DC,DS) WILL
|
||||
! NEXT BE CALLED TO APPLY THE TRANSFORMATION TO A 2 BY N MATRIX.
|
||||
!
|
||||
! ------------------------------------------------------------------
|
||||
|
||||
REAL (dp), INTENT(IN OUT) :: da
|
||||
REAL (dp), INTENT(IN OUT) :: db
|
||||
REAL (dp), INTENT(OUT) :: dc
|
||||
REAL (dp), INTENT(OUT) :: ds
|
||||
|
||||
REAL (dp) :: u, v, r
|
||||
IF (ABS(da) <= ABS(db)) GO TO 10
|
||||
|
||||
! *** HERE ABS(DA) > ABS(DB) ***
|
||||
|
||||
u = da + da
|
||||
v = db / u
|
||||
|
||||
! NOTE THAT U AND R HAVE THE SIGN OF DA
|
||||
|
||||
r = SQRT(.25D0 + v**2) * u
|
||||
|
||||
! NOTE THAT DC IS POSITIVE
|
||||
|
||||
dc = da / r
|
||||
ds = v * (dc + dc)
|
||||
db = ds
|
||||
da = r
|
||||
RETURN
|
||||
|
||||
! *** HERE ABS(DA) <= ABS(DB) ***
|
||||
|
||||
10 IF (db == 0.d0) GO TO 20
|
||||
u = db + db
|
||||
v = da / u
|
||||
|
||||
! NOTE THAT U AND R HAVE THE SIGN OF DB
|
||||
! (R IS IMMEDIATELY STORED IN DA)
|
||||
|
||||
da = SQRT(.25D0 + v**2) * u
|
||||
|
||||
! NOTE THAT DS IS POSITIVE
|
||||
|
||||
ds = db / da
|
||||
dc = v * (ds + ds)
|
||||
IF (dc == 0.d0) GO TO 15
|
||||
db = 1.d0 / dc
|
||||
RETURN
|
||||
15 db = 1.d0
|
||||
RETURN
|
||||
|
||||
! *** HERE DA = DB = 0.D0 ***
|
||||
|
||||
20 dc = 1.d0
|
||||
ds = 0.d0
|
||||
RETURN
|
||||
|
||||
END SUBROUTINE drotg
|
||||
|
||||
|
||||
SUBROUTINE dswap1 (n, dx, dy)
|
||||
! INTERCHANGES TWO VECTORS.
|
||||
! USES UNROLLED LOOPS FOR INCREMENTS EQUAL ONE.
|
||||
! JACK DONGARRA, LINPACK, 3/11/78.
|
||||
! This version is for increments = 1.
|
||||
|
||||
INTEGER, INTENT(IN) :: n
|
||||
REAL (dp), INTENT(IN OUT) :: dx(*)
|
||||
REAL (dp), INTENT(IN OUT) :: dy(*)
|
||||
|
||||
REAL (dp) :: dtemp
|
||||
INTEGER :: i, m, mp1
|
||||
|
||||
IF(n <= 0) RETURN
|
||||
|
||||
! CODE FOR BOTH INCREMENTS EQUAL TO 1
|
||||
!
|
||||
! CLEAN-UP LOOP
|
||||
|
||||
m = MOD(n,3)
|
||||
IF( m == 0 ) GO TO 40
|
||||
DO i = 1,m
|
||||
dtemp = dx(i)
|
||||
dx(i) = dy(i)
|
||||
dy(i) = dtemp
|
||||
END DO
|
||||
IF( n < 3 ) RETURN
|
||||
40 mp1 = m + 1
|
||||
DO i = mp1,n,3
|
||||
dtemp = dx(i)
|
||||
dx(i) = dy(i)
|
||||
dy(i) = dtemp
|
||||
dtemp = dx(i + 1)
|
||||
dx(i + 1) = dy(i + 1)
|
||||
dy(i + 1) = dtemp
|
||||
dtemp = dx(i + 2)
|
||||
dx(i + 2) = dy(i + 2)
|
||||
dy(i + 2) = dtemp
|
||||
END DO
|
||||
RETURN
|
||||
END SUBROUTINE dswap1
|
||||
|
||||
|
||||
SUBROUTINE drot1 (n, dx, dy, c, s)
|
||||
! APPLIES A PLANE ROTATION.
|
||||
! JACK DONGARRA, LINPACK, 3/11/78.
|
||||
! This version is for increments = 1.
|
||||
|
||||
INTEGER, INTENT(IN) :: n
|
||||
REAL (dp), INTENT(IN OUT) :: dx(*)
|
||||
REAL (dp), INTENT(IN OUT) :: dy(*)
|
||||
REAL (dp), INTENT(IN) :: c
|
||||
REAL (dp), INTENT(IN) :: s
|
||||
|
||||
REAL (dp) :: dtemp
|
||||
INTEGER :: i
|
||||
|
||||
IF(n <= 0) RETURN
|
||||
! CODE FOR BOTH INCREMENTS EQUAL TO 1
|
||||
|
||||
DO i = 1,n
|
||||
dtemp = c*dx(i) + s*dy(i)
|
||||
dy(i) = c*dy(i) - s*dx(i)
|
||||
dx(i) = dtemp
|
||||
END DO
|
||||
RETURN
|
||||
END SUBROUTINE drot1
|
||||
|
||||
|
||||
SUBROUTINE dsvdc(x, n, p, s, e, u, v, job, info)
|
||||
|
||||
INTEGER, INTENT(IN) :: n
|
||||
INTEGER, INTENT(IN) :: p
|
||||
REAL (dp), INTENT(IN OUT) :: x(:,:)
|
||||
REAL (dp), INTENT(OUT) :: s(:)
|
||||
REAL (dp), INTENT(OUT) :: e(:)
|
||||
REAL (dp), INTENT(OUT) :: u(:,:)
|
||||
REAL (dp), INTENT(OUT) :: v(:,:)
|
||||
INTEGER, INTENT(IN) :: job
|
||||
INTEGER, INTENT(OUT) :: info
|
||||
|
||||
! DSVDC IS A SUBROUTINE TO REDUCE A DOUBLE PRECISION NXP MATRIX X
|
||||
! BY ORTHOGONAL TRANSFORMATIONS U AND V TO DIAGONAL FORM. THE
|
||||
! DIAGONAL ELEMENTS S(I) ARE THE SINGULAR VALUES OF X. THE
|
||||
! COLUMNS OF U ARE THE CORRESPONDING LEFT SINGULAR VECTORS,
|
||||
! AND THE COLUMNS OF V THE RIGHT SINGULAR VECTORS.
|
||||
!
|
||||
! ON ENTRY
|
||||
!
|
||||
! X DOUBLE PRECISION(LDX,P), WHERE LDX.GE.N.
|
||||
! X CONTAINS THE MATRIX WHOSE SINGULAR VALUE
|
||||
! DECOMPOSITION IS TO BE COMPUTED. X IS
|
||||
! DESTROYED BY DSVDC.
|
||||
!
|
||||
! LDX INTEGER.
|
||||
! LDX IS THE LEADING DIMENSION OF THE ARRAY X.
|
||||
!
|
||||
! N INTEGER.
|
||||
! N IS THE NUMBER OF ROWS OF THE MATRIX X.
|
||||
!
|
||||
! P INTEGER.
|
||||
! P IS THE NUMBER OF COLUMNS OF THE MATRIX X.
|
||||
!
|
||||
! LDU INTEGER.
|
||||
! LDU IS THE LEADING DIMENSION OF THE ARRAY U.
|
||||
! (SEE BELOW).
|
||||
!
|
||||
! LDV INTEGER.
|
||||
! LDV IS THE LEADING DIMENSION OF THE ARRAY V.
|
||||
! (SEE BELOW).
|
||||
!
|
||||
! JOB INTEGER.
|
||||
! JOB CONTROLS THE COMPUTATION OF THE SINGULAR
|
||||
! VECTORS. IT HAS THE DECIMAL EXPANSION AB
|
||||
! WITH THE FOLLOWING MEANING
|
||||
!
|
||||
! A.EQ.0 DO NOT COMPUTE THE LEFT SINGULAR VECTORS.
|
||||
! A.EQ.1 RETURN THE N LEFT SINGULAR VECTORS IN U.
|
||||
! A.GE.2 RETURN THE FIRST MIN(N,P) SINGULAR
|
||||
! VECTORS IN U.
|
||||
! B.EQ.0 DO NOT COMPUTE THE RIGHT SINGULAR VECTORS.
|
||||
! B.EQ.1 RETURN THE RIGHT SINGULAR VECTORS IN V.
|
||||
!
|
||||
! ON RETURN
|
||||
!
|
||||
! S DOUBLE PRECISION(MM), WHERE MM=MIN(N+1,P).
|
||||
! THE FIRST MIN(N,P) ENTRIES OF S CONTAIN THE SINGULAR
|
||||
! VALUES OF X ARRANGED IN DESCENDING ORDER OF MAGNITUDE.
|
||||
!
|
||||
! E DOUBLE PRECISION(P).
|
||||
! E ORDINARILY CONTAINS ZEROS. HOWEVER SEE THE
|
||||
! DISCUSSION OF INFO FOR EXCEPTIONS.
|
||||
!
|
||||
! U DOUBLE PRECISION(LDU,K), WHERE LDU.GE.N. IF
|
||||
! JOBA.EQ.1 THEN K.EQ.N, IF JOBA.GE.2
|
||||
! THEN K.EQ.MIN(N,P).
|
||||
! U CONTAINS THE MATRIX OF LEFT SINGULAR VECTORS.
|
||||
! U IS NOT REFERENCED IF JOBA.EQ.0. IF N.LE.P
|
||||
! OR IF JOBA.EQ.2, THEN U MAY BE IDENTIFIED WITH X
|
||||
! IN THE SUBROUTINE CALL.
|
||||
!
|
||||
! V DOUBLE PRECISION(LDV,P), WHERE LDV.GE.P.
|
||||
! V CONTAINS THE MATRIX OF RIGHT SINGULAR VECTORS.
|
||||
! V IS NOT REFERENCED IF JOB.EQ.0. IF P.LE.N,
|
||||
! THEN V MAY BE IDENTIFIED WITH X IN THE
|
||||
! SUBROUTINE CALL.
|
||||
!
|
||||
! INFO INTEGER.
|
||||
! THE SINGULAR VALUES (AND THEIR CORRESPONDING SINGULAR
|
||||
! VECTORS) S(INFO+1),S(INFO+2),...,S(M) ARE CORRECT
|
||||
! (HERE M=MIN(N,P)). THUS IF INFO.EQ.0, ALL THE
|
||||
! SINGULAR VALUES AND THEIR VECTORS ARE CORRECT.
|
||||
! IN ANY EVENT, THE MATRIX B = TRANS(U)*X*V IS THE
|
||||
! BIDIAGONAL MATRIX WITH THE ELEMENTS OF S ON ITS DIAGONAL
|
||||
! AND THE ELEMENTS OF E ON ITS SUPER-DIAGONAL (TRANS(U)
|
||||
! IS THE TRANSPOSE OF U). THUS THE SINGULAR VALUES
|
||||
! OF X AND B ARE THE SAME.
|
||||
!
|
||||
! LINPACK. THIS VERSION DATED 03/19/79 .
|
||||
! G.W. STEWART, UNIVERSITY OF MARYLAND, ARGONNE NATIONAL LAB.
|
||||
!
|
||||
! DSVDC USES THE FOLLOWING FUNCTIONS AND SUBPROGRAMS.
|
||||
!
|
||||
! EXTERNAL DROT
|
||||
! BLAS DAXPY,DDOT,DSCAL,DSWAP,DNRM2,DROTG
|
||||
! FORTRAN DABS,DMAX1,MAX0,MIN0,MOD,DSQRT
|
||||
|
||||
! INTERNAL VARIABLES
|
||||
|
||||
INTEGER :: iter, j, jobu, k, kase, kk, l, ll, lls, lm1, lp1, ls,
|
||||
& lu, m, maxit,mm, mm1, mp1, nct, nctp1, ncu, nrt, nrtp1
|
||||
REAL (dp) :: t, work(n)
|
||||
REAL (dp) :: b, c, cs, el, emm1, f, g, scale, shift, sl, sm, sn,
|
||||
& smm1, t1, test, ztest
|
||||
LOGICAL :: wantu, wantv
|
||||
|
||||
! SET THE MAXIMUM NUMBER OF ITERATIONS.
|
||||
|
||||
maxit = 30
|
||||
|
||||
! DETERMINE WHAT IS TO BE COMPUTED.
|
||||
|
||||
wantu = .false.
|
||||
wantv = .false.
|
||||
jobu = MOD(job,100)/10
|
||||
ncu = n
|
||||
IF (jobu > 1) ncu = MIN(n,p)
|
||||
IF (jobu /= 0) wantu = .true.
|
||||
IF (MOD(job,10) /= 0) wantv = .true.
|
||||
|
||||
! REDUCE X TO BIDIAGONAL FORM, STORING THE DIAGONAL ELEMENTS
|
||||
! IN S AND THE SUPER-DIAGONAL ELEMENTS IN E.
|
||||
|
||||
info = 0
|
||||
nct = MIN(n-1, p)
|
||||
s(1:nct+1) = 0.0_dp
|
||||
nrt = MAX(0, MIN(p-2,n))
|
||||
lu = MAX(nct,nrt)
|
||||
IF (lu < 1) GO TO 170
|
||||
DO l = 1, lu
|
||||
lp1 = l + 1
|
||||
IF (l > nct) GO TO 20
|
||||
|
||||
! COMPUTE THE TRANSFORMATION FOR THE L-TH COLUMN AND
|
||||
! PLACE THE L-TH DIAGONAL IN S(L).
|
||||
|
||||
s(l) = SQRT( SUM( x(l:n,l)**2 ) )
|
||||
IF (s(l) == 0.0D0) GO TO 10
|
||||
IF (x(l,l) /= 0.0D0) s(l) = SIGN(s(l), x(l,l))
|
||||
x(l:n,l) = x(l:n,l) / s(l)
|
||||
x(l,l) = 1.0D0 + x(l,l)
|
||||
|
||||
10 s(l) = -s(l)
|
||||
|
||||
20 IF (p < lp1) GO TO 50
|
||||
DO j = lp1, p
|
||||
IF (l > nct) GO TO 30
|
||||
IF (s(l) == 0.0D0) GO TO 30
|
||||
|
||||
! APPLY THE TRANSFORMATION.
|
||||
|
||||
t = -DOT_PRODUCT(x(l:n,l), x(l:n,j)) / x(l,l)
|
||||
x(l:n,j) = x(l:n,j) + t * x(l:n,l)
|
||||
|
||||
! PLACE THE L-TH ROW OF X INTO E FOR THE
|
||||
! SUBSEQUENT CALCULATION OF THE ROW TRANSFORMATION.
|
||||
|
||||
30 e(j) = x(l,j)
|
||||
END DO
|
||||
|
||||
50 IF (.NOT.wantu .OR. l > nct) GO TO 70
|
||||
|
||||
! PLACE THE TRANSFORMATION IN U FOR SUBSEQUENT BACK MULTIPLICATION.
|
||||
|
||||
u(l:n,l) = x(l:n,l)
|
||||
|
||||
70 IF (l > nrt) CYCLE
|
||||
|
||||
! COMPUTE THE L-TH ROW TRANSFORMATION AND PLACE THE
|
||||
! L-TH SUPER-DIAGONAL IN E(L).
|
||||
|
||||
e(l) = SQRT( SUM( e(lp1:p)**2 ) )
|
||||
IF (e(l) == 0.0D0) GO TO 80
|
||||
IF (e(lp1) /= 0.0D0) e(l) = SIGN(e(l), e(lp1))
|
||||
e(lp1:lp1+p-l-1) = e(lp1:p) / e(l)
|
||||
e(lp1) = 1.0D0 + e(lp1)
|
||||
|
||||
80 e(l) = -e(l)
|
||||
IF (lp1 > n .OR. e(l) == 0.0D0) GO TO 120
|
||||
|
||||
! APPLY THE TRANSFORMATION.
|
||||
|
||||
work(lp1:n) = 0.0D0
|
||||
DO j = lp1, p
|
||||
work(lp1:lp1+n-l-1) = work(lp1:lp1+n-l-1) + e(j) *
|
||||
& x(lp1:lp1+n-l-1,j)
|
||||
END DO
|
||||
DO j = lp1, p
|
||||
x(lp1:lp1+n-l-1,j) = x(lp1:lp1+n-l-1,j) - (e(j)/e(lp1)) *
|
||||
& work(lp1:lp1+n-l-1)
|
||||
END DO
|
||||
|
||||
120 IF (.NOT.wantv) CYCLE
|
||||
|
||||
! PLACE THE TRANSFORMATION IN V FOR SUBSEQUENT
|
||||
! BACK MULTIPLICATION.
|
||||
|
||||
v(lp1:p,l) = e(lp1:p)
|
||||
END DO
|
||||
|
||||
! SET UP THE FINAL BIDIAGONAL MATRIX OF ORDER M.
|
||||
|
||||
170 m = MIN(p,n+1)
|
||||
nctp1 = nct + 1
|
||||
nrtp1 = nrt + 1
|
||||
IF (nct < p) s(nctp1) = x(nctp1,nctp1)
|
||||
IF (n < m) s(m) = 0.0D0
|
||||
IF (nrtp1 < m) e(nrtp1) = x(nrtp1,m)
|
||||
e(m) = 0.0D0
|
||||
|
||||
! IF REQUIRED, GENERATE U.
|
||||
|
||||
IF (.NOT.wantu) GO TO 300
|
||||
IF (ncu < nctp1) GO TO 200
|
||||
DO j = nctp1, ncu
|
||||
u(1:n,j) = 0.0_dp
|
||||
u(j,j) = 1.0_dp
|
||||
END DO
|
||||
|
||||
200 DO ll = 1, nct
|
||||
l = nct - ll + 1
|
||||
IF (s(l) == 0.0D0) GO TO 250
|
||||
lp1 = l + 1
|
||||
IF (ncu < lp1) GO TO 220
|
||||
DO j = lp1, ncu
|
||||
t = -DOT_PRODUCT(u(l:n,l), u(l:n,j)) / u(l,l)
|
||||
u(l:n,j) = u(l:n,j) + t * u(l:n,l)
|
||||
END DO
|
||||
|
||||
220 u(l:n,l) = -u(l:n,l)
|
||||
u(l,l) = 1.0D0 + u(l,l)
|
||||
lm1 = l - 1
|
||||
IF (lm1 < 1) CYCLE
|
||||
u(1:lm1,l) = 0.0_dp
|
||||
CYCLE
|
||||
|
||||
250 u(1:n,l) = 0.0_dp
|
||||
u(l,l) = 1.0_dp
|
||||
END DO
|
||||
|
||||
! IF IT IS REQUIRED, GENERATE V.
|
||||
|
||||
300 IF (.NOT.wantv) GO TO 350
|
||||
DO ll = 1, p
|
||||
l = p - ll + 1
|
||||
lp1 = l + 1
|
||||
IF (l > nrt) GO TO 320
|
||||
IF (e(l) == 0.0D0) GO TO 320
|
||||
DO j = lp1, p
|
||||
t = -DOT_PRODUCT(v(lp1:lp1+p-l-1,l),
|
||||
& v(lp1:lp1+p-l-1,j)) / v(lp1,l)
|
||||
v(lp1:lp1+p-l-1,j) = v(lp1:lp1+p-l-1,j) + t * v(lp1:lp1+p-l-1,l)
|
||||
END DO
|
||||
|
||||
320 v(1:p,l) = 0.0D0
|
||||
v(l,l) = 1.0D0
|
||||
END DO
|
||||
|
||||
! MAIN ITERATION LOOP FOR THE SINGULAR VALUES.
|
||||
|
||||
350 mm = m
|
||||
iter = 0
|
||||
|
||||
! QUIT IF ALL THE SINGULAR VALUES HAVE BEEN FOUND.
|
||||
|
||||
! ...EXIT
|
||||
360 IF (m == 0) GO TO 620
|
||||
|
||||
! IF TOO MANY ITERATIONS HAVE BEEN PERFORMED, SET FLAG AND RETURN.
|
||||
|
||||
IF (iter < maxit) GO TO 370
|
||||
info = m
|
||||
! ......EXIT
|
||||
GO TO 620
|
||||
|
||||
! THIS SECTION OF THE PROGRAM INSPECTS FOR NEGLIGIBLE ELEMENTS
|
||||
! IN THE S AND E ARRAYS. ON COMPLETION
|
||||
! THE VARIABLES KASE AND L ARE SET AS FOLLOWS.
|
||||
!
|
||||
! KASE = 1 IF S(M) AND E(L-1) ARE NEGLIGIBLE AND L < M
|
||||
! KASE = 2 IF S(L) IS NEGLIGIBLE AND L < M
|
||||
! KASE = 3 IF E(L-1) IS NEGLIGIBLE, L < M, AND
|
||||
! S(L), ..., S(M) ARE NOT NEGLIGIBLE (QR STEP).
|
||||
! KASE = 4 IF E(M-1) IS NEGLIGIBLE (CONVERGENCE).
|
||||
|
||||
370 DO ll = 1, m
|
||||
l = m - ll
|
||||
! ...EXIT
|
||||
IF (l == 0) EXIT
|
||||
test = ABS(s(l)) + ABS(s(l+1))
|
||||
ztest = test + ABS(e(l))
|
||||
IF (ztest /= test) CYCLE
|
||||
e(l) = 0.0D0
|
||||
! ......EXIT
|
||||
EXIT
|
||||
END DO
|
||||
|
||||
IF (l /= m - 1) GO TO 410
|
||||
kase = 4
|
||||
GO TO 480
|
||||
|
||||
410 lp1 = l + 1
|
||||
mp1 = m + 1
|
||||
DO lls = lp1, mp1
|
||||
ls = m - lls + lp1
|
||||
! ...EXIT
|
||||
IF (ls == l) EXIT
|
||||
test = 0.0D0
|
||||
IF (ls /= m) test = test + ABS(e(ls))
|
||||
IF (ls /= l + 1) test = test + ABS(e(ls-1))
|
||||
ztest = test + ABS(s(ls))
|
||||
IF (ztest /= test) CYCLE
|
||||
s(ls) = 0.0D0
|
||||
! ......EXIT
|
||||
EXIT
|
||||
END DO
|
||||
|
||||
IF (ls /= l) GO TO 450
|
||||
kase = 3
|
||||
GO TO 480
|
||||
|
||||
450 IF (ls /= m) GO TO 460
|
||||
kase = 1
|
||||
GO TO 480
|
||||
|
||||
460 kase = 2
|
||||
l = ls
|
||||
480 l = l + 1
|
||||
|
||||
! PERFORM THE TASK INDICATED BY KASE.
|
||||
|
||||
SELECT CASE ( kase )
|
||||
CASE ( 1)
|
||||
GO TO 490
|
||||
CASE ( 2)
|
||||
GO TO 520
|
||||
CASE ( 3)
|
||||
GO TO 540
|
||||
CASE ( 4)
|
||||
GO TO 570
|
||||
END SELECT
|
||||
|
||||
! DEFLATE NEGLIGIBLE S(M).
|
||||
|
||||
490 mm1 = m - 1
|
||||
f = e(m-1)
|
||||
e(m-1) = 0.0D0
|
||||
DO kk = l, mm1
|
||||
k = mm1 - kk + l
|
||||
t1 = s(k)
|
||||
CALL drotg(t1, f, cs, sn)
|
||||
s(k) = t1
|
||||
IF (k == l) GO TO 500
|
||||
f = -sn*e(k-1)
|
||||
e(k-1) = cs*e(k-1)
|
||||
|
||||
500 IF (wantv) CALL drot1(p, v(1:,k), v(1:,m), cs, sn)
|
||||
END DO
|
||||
GO TO 610
|
||||
|
||||
! SPLIT AT NEGLIGIBLE S(L).
|
||||
|
||||
520 f = e(l-1)
|
||||
e(l-1) = 0.0D0
|
||||
DO k = l, m
|
||||
t1 = s(k)
|
||||
CALL drotg(t1, f, cs, sn)
|
||||
s(k) = t1
|
||||
f = -sn*e(k)
|
||||
e(k) = cs*e(k)
|
||||
IF (wantu) CALL drot1(n, u(1:,k), u(1:,l-1), cs, sn)
|
||||
END DO
|
||||
GO TO 610
|
||||
|
||||
! PERFORM ONE QR STEP.
|
||||
!
|
||||
! CALCULATE THE SHIFT.
|
||||
|
||||
540 scale = MAX(ABS(s(m)),ABS(s(m-1)),ABS(e(m-1)),ABS(s(l)),ABS(e(l)))
|
||||
sm = s(m)/scale
|
||||
smm1 = s(m-1)/scale
|
||||
emm1 = e(m-1)/scale
|
||||
sl = s(l)/scale
|
||||
el = e(l)/scale
|
||||
b = ((smm1 + sm)*(smm1 - sm) + emm1**2)/2.0D0
|
||||
c = (sm*emm1)**2
|
||||
shift = 0.0D0
|
||||
IF (b == 0.0D0 .AND. c == 0.0D0) GO TO 550
|
||||
shift = SQRT(b**2+c)
|
||||
IF (b < 0.0D0) shift = -shift
|
||||
shift = c/(b + shift)
|
||||
|
||||
550 f = (sl + sm)*(sl - sm) - shift
|
||||
g = sl*el
|
||||
|
||||
! CHASE ZEROS.
|
||||
|
||||
mm1 = m - 1
|
||||
DO k = l, mm1
|
||||
CALL drotg(f, g, cs, sn)
|
||||
IF (k /= l) e(k-1) = f
|
||||
f = cs*s(k) + sn*e(k)
|
||||
e(k) = cs*e(k) - sn*s(k)
|
||||
g = sn*s(k+1)
|
||||
s(k+1) = cs*s(k+1)
|
||||
IF (wantv) CALL drot1(p, v(1:,k), v(1:,k+1), cs, sn)
|
||||
CALL drotg(f, g, cs, sn)
|
||||
s(k) = f
|
||||
f = cs*e(k) + sn*s(k+1)
|
||||
s(k+1) = -sn*e(k) + cs*s(k+1)
|
||||
g = sn*e(k+1)
|
||||
e(k+1) = cs*e(k+1)
|
||||
IF (wantu .AND. k < n) CALL drot1(n, u(1:,k), u(1:,k+1), cs, sn)
|
||||
END DO
|
||||
e(m-1) = f
|
||||
iter = iter + 1
|
||||
GO TO 610
|
||||
|
||||
! CONVERGENCE.
|
||||
|
||||
! MAKE THE SINGULAR VALUE POSITIVE.
|
||||
|
||||
570 IF (s(l) >= 0.0D0) GO TO 590
|
||||
s(l) = -s(l)
|
||||
IF (wantv) v(1:p,l) = -v(1:p,l)
|
||||
|
||||
! ORDER THE SINGULAR VALUE.
|
||||
|
||||
590 IF (l == mm) GO TO 600
|
||||
! ...EXIT
|
||||
IF (s(l) >= s(l+1)) GO TO 600
|
||||
t = s(l)
|
||||
s(l) = s(l+1)
|
||||
s(l+1) = t
|
||||
IF (wantv .AND. l < p) CALL dswap1(p, v(1:,l), v(1:,l+1))
|
||||
IF (wantu .AND. l < n) CALL dswap1(n, u(1:,l), u(1:,l+1))
|
||||
l = l + 1
|
||||
GO TO 590
|
||||
|
||||
600 iter = 0
|
||||
m = m - 1
|
||||
|
||||
610 GO TO 360
|
||||
|
||||
620 RETURN
|
||||
END SUBROUTINE dsvdc
|
||||
|
||||
END MODULE SVD
|
@ -1,189 +0,0 @@
|
||||
MODULE INTFCMOD
|
||||
IMPLICIT NONE
|
||||
PUBLIC :: INITLEVELS, TRANSF, COVG
|
||||
|
||||
CONTAINS
|
||||
SUBROUTINE INITLEVELS(T,HT,N,NU,Nv)
|
||||
USE TBRMOD
|
||||
USE SIZEMOD
|
||||
IMPLICIT NONE
|
||||
C INTEGER, PARAMETER:: NMAX = 101, RDIM = 10201
|
||||
C DIMENSION ULEVELS(1),Vlevels(1),T(1),HT(1),TG(1),XG(1),HH(101)
|
||||
REAL*8, DIMENSION(:), intent(in) :: T
|
||||
REAL*8, DIMENSION(:), intent(out) :: HT
|
||||
C INTEGER, intent(in) :: NG
|
||||
REAL*8 :: UMIN,UMAX,VMIN,VMAX, HU,HV
|
||||
integer :: N, I, NU, NV
|
||||
C REAL*8, DIMENSION(NMAX) :: HH
|
||||
C COMMON/TBR/HH
|
||||
|
||||
C IF (NG.GT.501) THEN
|
||||
C PRINT *,'Vector defining transformation of data > 501, stop'
|
||||
C STOP
|
||||
C END IF
|
||||
|
||||
|
||||
IF(N.ge.NMAX) then
|
||||
print *,'The number of wavelength points >',NMAX-1, ' stop'
|
||||
stop
|
||||
end if
|
||||
IF(N.lt.2) then
|
||||
print *,'The number of wavelength points < 2, stop'
|
||||
stop
|
||||
end if
|
||||
|
||||
HT(1)=0.5d0*(T(2)-T(1))
|
||||
HT(N)=0.5d0*(T(N)-T(N-1))
|
||||
HH(1)=-100.0d0
|
||||
HH(N)=-100.0d0
|
||||
DO I=2,N-1
|
||||
HT(I)=0.5d0*(T(I+1)-T(I-1))
|
||||
HH(I)=-100.0d0
|
||||
c10 CONTINUE
|
||||
enddo
|
||||
|
||||
|
||||
IF(NU.gt.NMAX) then
|
||||
print *,'The number of maxima >',NMAX,' stop'
|
||||
stop
|
||||
end if
|
||||
IF(NV.gt.NMAX) then
|
||||
print *,'The number of minima >',NMAX,' stop'
|
||||
stop
|
||||
end if
|
||||
|
||||
IF(NU.LT.1) Then
|
||||
print *,'The number of maxima < 1, stop'
|
||||
stop
|
||||
end if
|
||||
IF(NV.LT.1) Then
|
||||
print *,'The number of minima < 1, stop'
|
||||
stop
|
||||
end if
|
||||
|
||||
RETURN
|
||||
END SUBROUTINE INITLEVELS
|
||||
|
||||
|
||||
SUBROUTINE TRANSF(N,T,A,TIMEV,VALUE,DER)
|
||||
C
|
||||
C N number of data points
|
||||
C TIMEV vector of time points
|
||||
C A a vector of values of a function G(TIME)
|
||||
C T independent time point
|
||||
C VALUE is a value of a function at T, i.e. VALUE=G(T).
|
||||
c DER=G'(t)
|
||||
C
|
||||
USE SIZEMOD
|
||||
IMPLICIT NONE
|
||||
REAL*8, intent(inout):: VALUE, DER,T
|
||||
C INTEGER, PARAMETER :: RDIM = 10201
|
||||
REAL*8, DIMENSION(:), intent(in) :: A,TIMEV
|
||||
integer, intent(in) :: N
|
||||
REAL*8:: T1
|
||||
integer :: I
|
||||
|
||||
IF (T.LT.TIMEV(1)) then
|
||||
der=(A(2)-A(1))/(TIMEV(2)-TIMEV(1))
|
||||
T1=T-TIMEV(1)
|
||||
VALUE=A(1)+T1*DER
|
||||
return
|
||||
end if
|
||||
IF (T.GT.TIMEV(N)) then
|
||||
der = (A(N)-A(N-1))/(TIMEV(N)-TIMEV(N-1))
|
||||
T1 = T-TIMEV(N)
|
||||
VALUE=A(N)+T1*DER
|
||||
return
|
||||
end if
|
||||
DO I=2,N
|
||||
IF (T.LT.TIMEV(I)) GO TO 10
|
||||
ENDDO
|
||||
10 I=I-1
|
||||
T1=T-TIMEV(I)
|
||||
DER=(A(I+1)-A(I))/(TIMEV(i+1)-TIMEV(I))
|
||||
VALUE=A(I)+T1*DER
|
||||
RETURN
|
||||
END SUBROUTINE TRANSF
|
||||
|
||||
REAL*8 FUNCTION SPLE(N,T,A,TIMEV)
|
||||
C
|
||||
C N number of data points
|
||||
C TIME vector of time points
|
||||
C A a vector of values of a function G(TIME)
|
||||
C T independent time point
|
||||
C SPLE is a value of a function at T, i.e. SPLE=G(T).
|
||||
C
|
||||
USE SIZEMOD
|
||||
IMPLICIT NONE
|
||||
INTEGER, INTENT(IN):: N
|
||||
|
||||
REAL*8, INTENT(IN) :: T
|
||||
REAL*8, DIMENSION(:), INTENT(IN) :: A,TIMEV
|
||||
REAL*8 :: T1
|
||||
INTEGER :: I
|
||||
SPLE=-9.9d0
|
||||
IF (T.LT.TIMEV(1) .OR. T.GT.TIMEV(N)) RETURN
|
||||
DO I=2,N
|
||||
IF (T.LT.TIMEV(I)) GO TO 10
|
||||
ENDDO
|
||||
10 I=I-1
|
||||
T1=T-TIMEV(I)
|
||||
SPLE=A(I)+T1*(A(I+1)-A(I))/(TIMEV(i+1)-TIMEV(I))
|
||||
RETURN
|
||||
END FUNCTION SPLE
|
||||
|
||||
SUBROUTINE COVG(XL0,XL2,XL4,COV1,COV2,COV3,COV,T,N)
|
||||
C
|
||||
C Covariance function and its four derivatives for a vector T of length N
|
||||
C is assumed in a vector COV; COV(1,...,N,1)=r(T), COV(1,...,N, 2)=r'(T), etc.
|
||||
C The vector COV should be of the shape N x 5.
|
||||
C
|
||||
C COVG Returns:
|
||||
C XL0,XL2,XL4 - spectral moments.
|
||||
C
|
||||
C Covariance matrices COV1=r'(T-T), COV2=r''(T-T) and COV3=r'''(T-T)
|
||||
C Dimension of COV1, COV2 should be atleast N*N.
|
||||
C
|
||||
USE SIZEMOD
|
||||
! IMPLICIT NONE
|
||||
C INTEGER, PARAMETER:: NMAX = 101, RDIM = 10201
|
||||
REAL*8, PARAMETER:: ZERO = 0.0d0
|
||||
REAL*8, intent(inout) :: XL0,XL2,XL4
|
||||
REAL*8, DIMENSION(N,5), intent(in) :: COV
|
||||
REAL*8, DIMENSION(N), intent(in) :: T
|
||||
REAL*8, DIMENSION(RDIM), intent(inout) :: COV1,COV2,COV3
|
||||
INTEGER, intent(in) :: N
|
||||
integer :: I, J, II
|
||||
REAL*8 :: TT, T0
|
||||
C
|
||||
C COV(Y(T),Y(0)) = COV(:,1)
|
||||
C DERIVATIVE COV(Y(T),Y(0)) = COV(:,2)
|
||||
C 2-DERIVATIVE COV(Y(T),Y(0)) = COV(:,3)
|
||||
C 3-DERIVATIVE COV(Y(T),Y(0)) = COV(:,4)
|
||||
C 4-DERIVATIVE COV(Y(T),Y(0)) = COV(:,5)
|
||||
|
||||
XL0 = COV(1,1)
|
||||
XL2 = -COV(1,3)
|
||||
XL4 = COV(1,5)
|
||||
! XL0 = SPLE(NT, ZERO, COV(:,1), T)
|
||||
! XL2 = -SPLE(NT, ZERO, COV(:,3), T)
|
||||
! XL4 = SPLE(NT, ZERO, COV(:,5), T)
|
||||
|
||||
II=0
|
||||
DO I=1,N
|
||||
DO J=1,N
|
||||
II = II+1
|
||||
T0 = T(J)-T(I)
|
||||
TT = ABS(T0)
|
||||
COV1(II) = SPLE(N, TT, COV(:,2), T)
|
||||
COV2(II) = SPLE(N, TT, COV(:,3), T)
|
||||
COV3(II) = SPLE(N, TT, COV(:,4), T)
|
||||
IF (T0.LT.0.0d0) then
|
||||
COV1(II)=-COV1(II)
|
||||
COV3(II)=-COV3(II)
|
||||
endif
|
||||
enddo
|
||||
enddo
|
||||
RETURN
|
||||
END SUBROUTINE COVG
|
||||
END module intfcmod
|
File diff suppressed because it is too large
Load Diff
@ -1,20 +0,0 @@
|
||||
"""
|
||||
builds mvn.pyd
|
||||
"""
|
||||
import os
|
||||
import sys
|
||||
|
||||
|
||||
from wafo.f2py_tools import f2py_call_str
|
||||
|
||||
def compile_all():
|
||||
f2py_call = f2py_call_str()
|
||||
print '=' * 75
|
||||
print 'compiling mvn'
|
||||
print '=' * 75
|
||||
|
||||
os.system(f2py_call + ' mvn.pyf mvndst.f -c ')
|
||||
|
||||
|
||||
if __name__ == '__main__':
|
||||
compile_all()
|
@ -1,39 +0,0 @@
|
||||
! -*- f90 -*-
|
||||
! Note: the context of this file is case sensitive.
|
||||
|
||||
python module mvn ! in
|
||||
interface ! in :mvn
|
||||
subroutine mvnun(d,n,lower,upper,means,covar,maxpts,abseps,releps,value,inform) ! in :mvn:mvndst.f
|
||||
integer intent(hide) :: d=shape(means,0)
|
||||
integer intent(hide) :: n=shape(means,1)
|
||||
double precision dimension(d) :: lower
|
||||
double precision dimension(d) :: upper
|
||||
double precision dimension(d,n) :: means
|
||||
double precision dimension(d,d) :: covar
|
||||
integer intent(optional) :: maxpts=d*1000
|
||||
double precision intent(optional) :: abseps=1e-6
|
||||
double precision intent(optional) :: releps=1e-6
|
||||
double precision intent(out) :: value
|
||||
integer intent(out) :: inform
|
||||
end subroutine mvnun
|
||||
|
||||
subroutine mvndst(n,lower,upper,infin,correl,maxpts,abseps,releps,error,value,inform) ! in :mvn:mvndst.f
|
||||
integer intent(hide) :: n=len(lower)
|
||||
double precision dimension(n) :: lower
|
||||
double precision dimension(n) :: upper
|
||||
integer dimension(n) :: infin
|
||||
double precision dimension(n*(n-1)/2) :: correl
|
||||
integer intent(optional) :: maxpts=2000
|
||||
double precision intent(optional) :: abseps=1e-6
|
||||
double precision intent(optional) :: releps=1e-6
|
||||
double precision intent(out) :: error
|
||||
double precision intent(out) :: value
|
||||
integer intent(out) :: inform
|
||||
integer :: ivls
|
||||
common /dkblck/ ivls
|
||||
end subroutine mvndst
|
||||
end interface
|
||||
end python module mvn
|
||||
|
||||
! This file was auto-generated with f2py (version:2.39.235_1752).
|
||||
! See http://cens.ioc.ee/projects/f2py2e/
|
File diff suppressed because it is too large
Load Diff
Some files were not shown because too many files have changed in this diff Show More
Loading…
Reference in New Issue