CoastSat_WRL/functions/utils.py

# -*- coding: utf-8 -*-
"""
Created on Thu Mar  1 11:30:31 2018

@author: z5030440

Contains all the utilities, convenience functions and small functions that do simple things
"""

import matplotlib.pyplot as plt
import numpy as np
import pdb


def ecdf(x):
    """convenience function for computing the empirical CDF"""
    vals, counts = np.unique(x, return_counts=True)
    ecdf = np.cumsum(counts).astype(np.float64)
    ecdf /= ecdf[-1]
    return vals, ecdf

def intensity_histogram(image):
    """plots histogram and cumulative distribution of the pixel intensities in an image"""
    imSize = image.shape
    if len(imSize) == 2:
        im = image[:,:].reshape(imSize[0] * imSize[1])
        im = im[~np.isnan(im)] 
        fig, (ax1, ax2) = plt.subplots(2,1, sharex=True, figsize = (8,6))
        ax1.hist(im, bins=300)
        ax1.set_title('Probability density function')
        ax2.hist(im, bins=300, cumulative=True, histtype='step')
        ax2.set_title('Cumulative distribution')
        plt.show()
    
    else:       
        for i in range(imSize[2]):
            im = image[:,:,i].reshape(imSize[0] * imSize[1])
            im = im[~np.isnan(im)] 
            fig, (ax1, ax2) = plt.subplots(2,1, sharex=True, figsize = (8,6))
            ax1.hist(im, bins=300)
            ax1.set_title('Probability density function')
            ax2.hist(im, bins=300, cumulative=True, histtype='step')
            ax2.set_title('Cumulative distribution')
            plt.show()
            
def compare_images(im1, im2):
    """plots 2 images next to each other, sharing the axis"""
    plt.figure()
    ax1 = plt.subplot(121)
    plt.imshow(im1, cmap='gray')
    ax2 = plt.subplot(122, sharex=ax1, sharey=ax1)
    plt.imshow(im2, cmap='gray')
    plt.show()  
    
def find_indices(lst, condition):
    "imitation of MATLAB find function"
    return [i for i, elem in enumerate(lst) if condition(elem)]

def reject_outliers(data, m=2):
    "rejects outliers in a numpy array"
    return data[abs(data - np.mean(data)) < m * np.std(data)]