Ensure float precision is limited when outputting csv

Without this change, interim .csv files were being recorded with 9 decimal places, greatly increasing the file size. This keeps the number of decimals limited to a practical amount.

src/analysis/forecast_twl.py

@@ -422,6 +422,6 @@ def create_twl_forecast(
         profile_type=profile_type,
     )
 
-    df_twl.to_csv(output_file)
+    df_twl.to_csv(output_file, float_format="%.4f")
     logger.info("Saved to %s", output_file)
     logger.info("Done!")

src/analysis/forecasted_storm_impacts.py

@@ -140,6 +140,6 @@ def create_forecasted_impacts(profile_features_csv, forecasted_twl_csv, output_f
         right_on=["site_id"],
     )
 
-    df_forecasted_impacts.to_csv(output_file)
+    df_forecasted_impacts.to_csv(output_file, float_format="%.4f")
     logger.info("Saved to %s", output_file)
     logger.info("Done!")

src/data/parse_mat.py

@@ -467,8 +467,8 @@ def create_waves_csv(waves_mat, sites_csv, waves_output_file, sites_waves_output
     df_waves.set_index(["site_id", "datetime"], inplace=True)
     df_waves.sort_index(inplace=True)
     df_waves, df_sites_waves = split_site_wave_params(df_waves)
-    df_waves.to_csv(waves_output_file)
+    df_waves.to_csv(waves_output_file, float_format="%.4f")
-    df_sites_waves.to_csv(sites_waves_output_file)
+    df_sites_waves.to_csv(sites_waves_output_file, float_format="%.4f")
     logger.info("Created %s", waves_output_file)
     logger.info("Created %s", sites_waves_output_file)
 
@@ -523,9 +523,9 @@ def create_sites_and_profiles_csv(
     df_sites.set_index(["site_id"], inplace=True)
     df_sites.sort_index(inplace=True)
 
-    df_profiles.to_csv(profiles_output_file)
+    df_profiles.to_csv(profiles_output_file, float_format="%.8f")
     logger.info("Created %s", profiles_output_file)
-    df_sites.to_csv(sites_output_file, float_format="%.3f")
+    df_sites.to_csv(sites_output_file, float_format="%.8f")
     logger.info("Created %s", sites_output_file)
 
 
@@ -542,7 +542,7 @@ def create_tides_csv(tides_mat, sites_csv, output_file):
     df_tides = replace_unique_sites(df_tides, df_sites)
     df_tides.set_index(["site_id", "datetime"], inplace=True)
     df_tides.sort_index(inplace=True)
-    df_tides.to_csv(output_file)
+    df_tides.to_csv(output_file, float_format="%.4f")
     logger.info("Created %s", output_file)