main.py

import errno
from datetime import timedelta

import numpy as np

from utils.utils import delimit_data, resampling_data, \
    add_extra_features, plot_similarity_prediction, predict_all, get_module_logger, plot_events_icme,  \
    validate_time_format, get_or_download_data_set, convert_events_to_np_array
from utils.gautierfunctions.process import turn_peaks_to_clouds

import logging.config
import sys
import os

PLOT = False
"""
Plot results or not
"""

DEBUG = False
"""Run in debug mode without passing arguments to the module and using directly START_TIME_DEBUG and STOP_TIME_DEBUG 
defined just below """

START_TIME_DEBUG = "2012-7-2T20:30:00"
"""
The start time to use in order to debug, feel free to modify it
"""

STOP_TIME_DEBUG = "2012-7-20T23:59:00"
"""
The stop time to use in order to debug, feel free to modify it
"""

DELTA_TIME = 0
"""
Add a time to the start/stop times, by default is zero
"""


def main():
    """
    Returns the predicted ICME events for a given start and stop times,
    then saves them into csv file named 'icme_events.csv' and into
    image named 'icme_events.png'


    if you are in the debug mode (DEBUG == True) then you need to specify the start/stop times manually (see
    START_TIME_DEBUG and STOP_TIME_DEBUG), if you are in the production mode (DEBUG == False) then you need to pass
    start/stop times as arguments (see readme.md)

    if you need to plot the results you need to assign True to PLOT

    :return: the predicted ICME events
    """
    logger.info("Checking arguments... ")
    nb_arguments = len(sys.argv)
    destination_folder_path = ""
    if not DEBUG and nb_arguments < 3:
        message = "X X : Expected at least 3 arguments. You give only: ", nb_arguments, *sys.argv
        logger.error(message)
        sys.exit(errno.EINVAL)
    elif not DEBUG and nb_arguments == 4:
        message = "Destination folder is given: ", sys.argv[3]
        logger.info(message)
        destination_folder_path = sys.argv[3]
    elif not DEBUG and nb_arguments == 3:
        message = "Destination folder is missing, the output files will be saved in the current folder"
        logger.warn(message)

    if not DEBUG:
        message = "Checking the formats of start time and stop time..."
        logger.info(message)
        try:
            start = validate_time_format(sys.argv[1])
            stop = validate_time_format(sys.argv[2])
        except ValueError as error:
            logger.error('Caught this error: ' + repr(error))
            sys.exit(errno.EINVAL)
    else:
        start = validate_time_format(START_TIME_DEBUG)
        stop = validate_time_format(STOP_TIME_DEBUG)

    logger.info("Arguments are valid")

    if start >= stop:
        message = "X X : The start time must be less than the stop time: ", start, stop
        logger.error(message)
        sys.exit(errno.EINVAL)

    logger.info("Preparing data... ")
    raw_data = get_or_download_data_set()

    add_extra_features(raw_data)
    data = delimit_data(raw_data)
    data = resampling_data(data)

    prediction_start_time = start - timedelta(minutes=DELTA_TIME)
    prediction_stop_time = stop + timedelta(minutes=DELTA_TIME)

    data_prediction = data[data.index < prediction_stop_time]
    data_prediction = data_prediction[data_prediction.index > prediction_start_time]

    logger.info("Run predictions... ")
    prediction = predict_all(data_prediction)
    logger.info("Post processing... ")
    integral = prediction.sum(axis=1)
    list_icmes = turn_peaks_to_clouds(integral, 12)

    path_save = os.path.join(destination_folder_path, "similarity_output.png")
    plot_similarity_prediction(prediction, path_save, plot=PLOT)

    path_save = os.path.join(destination_folder_path, "events_output.png")
    plot_events_icme(integral, list_icmes, path_save, PLOT)

    logger.info("Writing results... ")
    events_catalog = convert_events_to_np_array(list_icmes)
    path_save = os.path.join(destination_folder_path, "events_tt_output.csv")
    np.savetxt(path_save, events_catalog, delimiter=" ", fmt='%s')
    message = "List of events %s ", *events_catalog
    logger.info(message)
    logger.info("End")


if __name__ == '__main__':
    logger = get_module_logger(__name__)
    # run predictions
    main()
    logging.shutdown()