gain_job.py

#
# Author: Axel Montout <axel.montout <a.t> bristol.ac.uk>
#
# Copyright (C) 2020  Biospi Laboratory for Medical Bioinformatics, University of Bristol, UK
#
# This file is part of PredictionOfHelminthsInfection.
#
# PHI is free software: you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation, either version 3 of the License, or
# (at your option) any later version.
#
# PHI is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
# GNU General Public License for more details.
#
# You should have received a copy of the GNU General Public License
# along with seaMass.  If not, see <http://www.gnu.org/licenses/>.
#

import argparse
import glob
import os
from pathlib import Path

import gain_imputation
from dataset import create_data_sets
from pipeline import add_exogenous
import ml
import numpy as np


def imputed_data_exists(path):
    folders = [x[0] for x in os.walk(path) if len(x) > 0 and "miss_rate" in x[0]]
    return len(folders) > 0, folders[0] if len(folders) > 0 else []


def famacha_dataset_exists(path):
    folders = [
        x[0]
        for x in os.walk(path)
        if len(x) > 0
        and "dataset" in x[0]
        and "temp" not in x[0]
        and "hum" not in x[0]
        and "temp_humidity" not in x[0]
    ]
    return (
        len(folders) > 0,
        [x for x in folders if "gain" in x and "entireday" in x][::-1],
        [x for x in folders if "li" in x and "entireday" in x][::-1],
        [x for x in folders if "gain" in x and "nightonly" in x][::-1],
        [x for x in folders if "li" in x and "nightonly" in x][::-1],
    )


def weather_files_exist(path):
    temperature_files = [
        glob.glob(x[0] + "/*.csv")[0]
        for x in os.walk(path)
        if len(x) > 0 and "temp" in x[0] and "temp_humidity" not in x[0]
    ]
    humidity_files = [
        glob.glob(x[0] + "/*.csv")[0]
        for x in os.walk(path)
        if len(x) > 0 and "hum" in x[0] and "temp_humidity" not in x[0]
    ]
    return (
        len(temperature_files) > 0 and len(humidity_files) > 0,
        temperature_files[::-1],
        humidity_files[::-1],
    )


if __name__ == "__main__":
    print("starting job...")
    print("imputation...")

    parser = argparse.ArgumentParser()
    parser.add_argument("data_dir", type=str)
    parser.add_argument("output_dir", type=str)
    parser.add_argument("fam_file", help="Famacha HDF5 file", type=str)
    parser.add_argument("weather_file", help="Weather file", type=str)

    parser.add_argument(
        "--batch_size",
        help="for imputation, the number of samples in mini-batch",
        default=128,
        type=int,
    )
    parser.add_argument(
        "--hint_rate", help="for imputation, hint probability", default=0.9, type=float
    )
    parser.add_argument(
        "--alpha", help="for imputation, hyperparameter", default=100, type=float
    )
    parser.add_argument(
        "--iterations",
        help="for imputation, number of training interations",
        default=600,
        type=int,
    )
    parser.add_argument(
        "--miss_rate",
        help="for imputation, missing data probability",
        default=0.0,
        type=float,
    )
    parser.add_argument("--n_job", type=int, default=6, help="Number of thread to use.")
    parser.add_argument(
        "--n_top_traces",
        type=int,
        default=-1,
        help="for imputation, select n traces with highest entropy (<= 0 to select all traces)",
    )
    parser.add_argument(
        "--enable_anscombe",
        help="for imputation, appy anscombe on activity count before imputation",
        type=bool,
        default=False,
    )
    parser.add_argument(
        "--enable_remove_zeros",
        help="for imputation, remove zero counts in activity before imputation",
        type=bool,
        default=False,
    )
    parser.add_argument(
        "--enable_log_anscombe",
        help="for imputation, appy log(anscombe) on activity count before imputation",
        type=bool,
        default=True,
    )
    parser.add_argument("--window", type=bool, default=False)
    parser.add_argument(
        "--export_csv",
        type=bool,
        help="for imputation, export imputed traces as csv",
        default=True,
    )
    parser.add_argument("--export_traces", type=bool, default=True)
    parser.add_argument(
        "--reshape",
        type=str,
        help="for imputation, reshape activity traces to 1 day chunck",
        default="y",
    )
    parser.add_argument("--w", type=str, default="n")
    parser.add_argument(
        "--add_t_col",
        help="for imputation, add time column in reshape",
        type=str,
        default="y",
    )
    parser.add_argument(
        "--thresh_daytime",
        help="for imputation, minimum number of positive count in 1 day.",
        default=100,
        type=int,
    )
    parser.add_argument(
        "--thresh_nan_ratio",
        help="for imputation, max percent of nan allowed in 1 day.",
        default=80,
        type=int,
    )

    parser.add_argument(
        "--ndays", help="Number of days in samples", default=7, type=int
    )

    args = parser.parse_args()

    exist, i_path = imputed_data_exists(args.output_dir)
    if not exist:
        imputed_data_dir = gain_imputation.start(args)
    else:
        print("imputation folder exists! Skipping to dataset creation...")
        imputed_data_dir = i_path
    # imputed_data_dir = "F:\Data2\job_debug\miss_rate_0_0_iteration_0600_thresh_100_anscombe_False_n_top_traces_17"
    print("imputation done.")

    (
        dataset_exist,
        dataset_files_gain,
        dataset_files_li,
        dataset_files_gain_night,
        dataset_files_li_night,
    ) = famacha_dataset_exists(args.output_dir)

    if not dataset_exist:
        famFile = Path(args.fam_file)
        dataDir = Path(imputed_data_dir)
        dataset_files_gain = []
        dataset_files_li = []
        dataset_files_gain_night = []
        dataset_files_li_night = []
        for night in [True, False]:

            night_label = "nightonly" if night else "entireday"

            if night:
                dataset_files_gain_night.append(
                    create_data_sets.main(
                        night,
                        famFile,
                        dataDir,
                        Path(args.output_dir + "/dataset_gain_7day_" + night_label),
                        "first_sensor_value_gain",
                        7,
                    )
                )
                dataset_files_gain_night.append(
                    create_data_sets.main(
                        night,
                        famFile,
                        dataDir,
                        Path(args.output_dir + "/dataset_gain_6day_" + night_label),
                        "first_sensor_value_gain",
                        6,
                    )
                )
                dataset_files_gain_night.append(
                    create_data_sets.main(
                        night,
                        famFile,
                        dataDir,
                        Path(args.output_dir + "/dataset_gain_5day_" + night_label),
                        "first_sensor_value_gain",
                        5,
                    )
                )
                dataset_files_gain_night.append(
                    create_data_sets.main(
                        night,
                        famFile,
                        dataDir,
                        Path(args.output_dir + "/dataset_gain_4day_" + night_label),
                        "first_sensor_value_gain",
                        4,
                    )
                )
                dataset_files_gain_night.append(
                    create_data_sets.main(
                        night,
                        famFile,
                        dataDir,
                        Path(args.output_dir + "/dataset_gain_3day_" + night_label),
                        "first_sensor_value_gain",
                        3,
                    )
                )
                dataset_files_gain_night.append(
                    create_data_sets.main(
                        night,
                        famFile,
                        dataDir,
                        Path(args.output_dir + "/dataset_gain_2day_" + night_label),
                        "first_sensor_value_gain",
                        2,
                    )
                )
                dataset_files_gain_night.append(
                    create_data_sets.main(
                        night,
                        famFile,
                        dataDir,
                        Path(args.output_dir + "/dataset_gain_1day_" + night_label),
                        "first_sensor_value_gain",
                        1,
                    )
                )

                dataset_files_li_night.append(
                    create_data_sets.main(
                        night,
                        famFile,
                        dataDir,
                        Path(args.output_dir + "/dataset_li_7day_" + night_label),
                        "first_sensor_value_li",
                        7,
                    )
                )
                dataset_files_li_night.append(
                    create_data_sets.main(
                        night,
                        famFile,
                        dataDir,
                        Path(args.output_dir + "/dataset_li_6day_" + night_label),
                        "first_sensor_value_li",
                        6,
                    )
                )
                dataset_files_li_night.append(
                    create_data_sets.main(
                        night,
                        famFile,
                        dataDir,
                        Path(args.output_dir + "/dataset_li_5day_" + night_label),
                        "first_sensor_value_li",
                        5,
                    )
                )
                dataset_files_li_night.append(
                    create_data_sets.main(
                        night,
                        famFile,
                        dataDir,
                        Path(args.output_dir + "/dataset_li_4day_" + night_label),
                        "first_sensor_value_li",
                        4,
                    )
                )
                dataset_files_li_night.append(
                    create_data_sets.main(
                        night,
                        famFile,
                        dataDir,
                        Path(args.output_dir + "/dataset_li_3day_" + night_label),
                        "first_sensor_value_li",
                        3,
                    )
                )
                dataset_files_li_night.append(
                    create_data_sets.main(
                        night,
                        famFile,
                        dataDir,
                        Path(args.output_dir + "/dataset_li_2day_" + night_label),
                        "first_sensor_value_li",
                        2,
                    )
                )
                dataset_files_li_night.append(
                    create_data_sets.main(
                        night,
                        famFile,
                        dataDir,
                        Path(args.output_dir + "/dataset_li_1day_" + night_label),
                        "first_sensor_value_li",
                        1,
                    )
                )
            else:
                dataset_files_gain.append(
                    create_data_sets.main(
                        night,
                        famFile,
                        dataDir,
                        Path(args.output_dir + "/dataset_gain_7day_" + night_label),
                        "first_sensor_value_gain",
                        7,
                    )
                )
                dataset_files_gain.append(
                    create_data_sets.main(
                        night,
                        famFile,
                        dataDir,
                        Path(args.output_dir + "/dataset_gain_6day_" + night_label),
                        "first_sensor_value_gain",
                        6,
                    )
                )
                dataset_files_gain.append(
                    create_data_sets.main(
                        night,
                        famFile,
                        dataDir,
                        Path(args.output_dir + "/dataset_gain_5day_" + night_label),
                        "first_sensor_value_gain",
                        5,
                    )
                )
                dataset_files_gain.append(
                    create_data_sets.main(
                        night,
                        famFile,
                        dataDir,
                        Path(args.output_dir + "/dataset_gain_4day_" + night_label),
                        "first_sensor_value_gain",
                        4,
                    )
                )
                dataset_files_gain.append(
                    create_data_sets.main(
                        night,
                        famFile,
                        dataDir,
                        Path(args.output_dir + "/dataset_gain_3day_" + night_label),
                        "first_sensor_value_gain",
                        3,
                    )
                )
                dataset_files_gain.append(
                    create_data_sets.main(
                        night,
                        famFile,
                        dataDir,
                        Path(args.output_dir + "/dataset_gain_2day_" + night_label),
                        "first_sensor_value_gain",
                        2,
                    )
                )
                dataset_files_gain.append(
                    create_data_sets.main(
                        night,
                        famFile,
                        dataDir,
                        Path(args.output_dir + "/dataset_gain_1day_" + night_label),
                        "first_sensor_value_gain",
                        1,
                    )
                )

                dataset_files_li.append(
                    create_data_sets.main(
                        night,
                        famFile,
                        dataDir,
                        Path(args.output_dir + "/dataset_li_7day_" + night_label),
                        "first_sensor_value_li",
                        7,
                    )
                )
                dataset_files_li.append(
                    create_data_sets.main(
                        night,
                        famFile,
                        dataDir,
                        Path(args.output_dir + "/dataset_li_6day_" + night_label),
                        "first_sensor_value_li",
                        6,
                    )
                )
                dataset_files_li.append(
                    create_data_sets.main(
                        night,
                        famFile,
                        dataDir,
                        Path(args.output_dir + "/dataset_li_5day_" + night_label),
                        "first_sensor_value_li",
                        5,
                    )
                )
                dataset_files_li.append(
                    create_data_sets.main(
                        night,
                        famFile,
                        dataDir,
                        Path(args.output_dir + "/dataset_li_4day_" + night_label),
                        "first_sensor_value_li",
                        4,
                    )
                )
                dataset_files_li.append(
                    create_data_sets.main(
                        night,
                        famFile,
                        dataDir,
                        Path(args.output_dir + "/dataset_li_3day_" + night_label),
                        "first_sensor_value_li",
                        3,
                    )
                )
                dataset_files_li.append(
                    create_data_sets.main(
                        night,
                        famFile,
                        dataDir,
                        Path(args.output_dir + "/dataset_li_2day_" + night_label),
                        "first_sensor_value_li",
                        2,
                    )
                )
                dataset_files_li.append(
                    create_data_sets.main(
                        night,
                        famFile,
                        dataDir,
                        Path(args.output_dir + "/dataset_li_1day_" + night_label),
                        "first_sensor_value_li",
                        1,
                    )
                )

    # dataset_files = ['F:\\Data2\\job_debug\\dataset_gain_7day', 'F:\\Data2\\job_debug\\dataset_gain_6day',
    #                  'F:\\Data2\\job_debug\\dataset_gain_5day', 'F:\\Data2\\job_debug\\dataset_gain_4day',
    #                  'F:\\Data2\\job_debug\\dataset_gain_3day', 'F:\\Data2\\job_debug\\dataset_gain_2day',
    #                  'F:\\Data2\\job_debug\\dataset_gain_1day']

    print(dataset_files_gain, dataset_files_li)
    print(dataset_files_gain_night, dataset_files_li_night)
    print("dataset creation done.")

    # dataset_exist, temperature_files, humidity_files = weather_files_exist(args.output_dir)
    # if not dataset_exist:
    #     print("create weather data.")
    #     humidity_files = []
    #     temperature_files = []
    #     for d_f in dataset_files_gain:
    #         files = glob.glob(d_f + "/*.csv")
    #         t_f, h_f = add_exogenous.main(files[0], d_f, args.weather_file)
    #         temperature_files.append(t_f)
    #         humidity_files.append(h_f)
    # print("weather data creation done.")
    humidity_files = [None, None, None, None, None, None, None]
    temperature_files = [None, None, None, None, None, None, None]

    steps = [
        # ["TEMPERATURE", "STANDARDSCALER"],
        # ["HUMIDITY", "STANDARDSCALER"],
        ["QN", "ANSCOMBE", "LOG", "DIFFLASTD"],
        ["QN", "ANSCOMBE", "LOG", "DIFFLASTD", "PCA(3)"],
        ["QN", "ANSCOMBE", "LOG"],
        ["QN", "ANSCOMBE", "LOG", "PCA(3)"],
        ["QN", "ANSCOMBE", "LOG", "DIFFLASTDAPPEND"],
        ["QN", "ANSCOMBE", "LOG", "DIFFLASTDAPPEND", "PCA(3)"],
        ["QN", "ANSCOMBE", "LOG", "DIFFLASTDAPPEND", "STANDARDSCALER"],
        ["QN", "ANSCOMBE", "LOG", "DIFFLASTDAPPEND", "STANDARDSCALER", "PCA(3)"],
        ["QN", "ANSCOMBE", "LOG", "DIFFAPPEND"],
        ["QN", "ANSCOMBE", "LOG", "DIFFAPPEND", "PCA(3)"],
        ["QN", "ANSCOMBE", "LOG", "DIFFAPPEND", "STANDARDSCALER"],
        ["QN", "ANSCOMBE", "LOG", "DIFFAPPEND", "STANDARDSCALER", "PCA(3)"],
        ["QN", "ANSCOMBE", "LOG", "DIFF"],
        ["QN", "ANSCOMBE", "LOG", "DIFF", "PCA(3)"],
        ["QN", "ANSCOMBE", "LOG", "CENTER", "CWT(MEXH)", "STANDARDSCALER"],
        ["QN", "ANSCOMBE", "LOG", "CENTER", "CWT(MORL)", "STANDARDSCALER"],
    ]

    steps_night = [
        ["QN", "ANSCOMBE", "LOG"],
        ["QN", "ANSCOMBE", "LOG", "PCA(3)"],
        ["QN", "ANSCOMBE", "LOG", "CENTER", "CWT(MEXH)", "STANDARDSCALER"],
        ["QN", "ANSCOMBE", "LOG", "CENTER", "CWT(MORL)", "STANDARDSCALER"],
    ]

    ml.main(
        steps_night,
        args.output_dir + "/ml/ml_kfold_2to2_7day_night",
        dataset_files_gain_night[0],
        "1To1",
        "2To2",
        False,
        12,
        humidity_files[0],
        temperature_files[0],
        5,
        10,
        20,
        6,
        True,
        True,
        "RepeatedKFold",
        6,
        60,
    )
    ml.main(
        steps_night,
        args.output_dir + "/ml/ml_kfold_2to2_6day_night",
        dataset_files_gain_night[1],
        "1To1",
        "2To2",
        False,
        12,
        humidity_files[1],
        temperature_files[1],
        5,
        10,
        20,
        6,
        True,
        True,
        "RepeatedKFold",
        6,
        60,
    )
    ml.main(
        steps_night,
        args.output_dir + "/ml/ml_kfold_2to2_5day_night",
        dataset_files_gain_night[2],
        "1To1",
        "2To2",
        False,
        12,
        humidity_files[2],
        temperature_files[2],
        5,
        10,
        20,
        6,
        True,
        True,
        "RepeatedKFold",
        6,
        60,
    )
    ml.main(
        steps_night,
        args.output_dir + "/ml/ml_kfold_2to2_4day_night",
        dataset_files_gain_night[3],
        "1To1",
        "2To2",
        False,
        12,
        humidity_files[3],
        temperature_files[3],
        5,
        10,
        20,
        6,
        True,
        True,
        "RepeatedKFold",
        6,
        60,
    )
    ml.main(
        steps_night,
        args.output_dir + "/ml/ml_kfold_2to2_3day_night",
        dataset_files_gain_night[4],
        "1To1",
        "2To2",
        False,
        12,
        humidity_files[4],
        temperature_files[4],
        5,
        10,
        20,
        6,
        True,
        True,
        "RepeatedKFold",
        6,
        60,
    )
    ml.main(
        steps_night,
        args.output_dir + "/ml/ml_kfold_2to2_2day_night",
        dataset_files_gain_night[5],
        "1To1",
        "2To2",
        False,
        12,
        humidity_files[5],
        temperature_files[5],
        5,
        10,
        20,
        6,
        True,
        True,
        "RepeatedKFold",
        6,
        60,
    )
    ml.main(
        steps_night,
        args.output_dir + "/ml/ml_kfold_2to2_1day_night",
        dataset_files_gain_night[6],
        "1To1",
        "2To2",
        False,
        12,
        humidity_files[6],
        temperature_files[6],
        5,
        10,
        20,
        6,
        True,
        True,
        "RepeatedKFold",
        6,
        60,
    )

    ml.main(
        steps,
        args.output_dir + "/ml/ml_kfold_2to2_7day",
        dataset_files_gain[0],
        "1To1",
        "2To2",
        False,
        12,
        humidity_files[0],
        temperature_files[0],
        5,
        10,
        20,
        6,
        True,
        True,
        "RepeatedKFold",
        6,
        60,
    )
    ml.main(
        steps,
        args.output_dir + "/ml/ml_kfold_2to2_6day",
        dataset_files_gain[1],
        "1To1",
        "2To2",
        False,
        12,
        humidity_files[1],
        temperature_files[1],
        5,
        10,
        20,
        6,
        True,
        True,
        "RepeatedKFold",
        6,
        60,
    )
    ml.main(
        steps,
        args.output_dir + "/ml/ml_kfold_2to2_5day",
        dataset_files_gain[2],
        "1To1",
        "2To2",
        False,
        12,
        humidity_files[2],
        temperature_files[2],
        5,
        10,
        20,
        6,
        True,
        True,
        "RepeatedKFold",
        6,
        60,
    )
    ml.main(
        steps,
        args.output_dir + "/ml/ml_kfold_2to2_4day",
        dataset_files_gain[3],
        "1To1",
        "2To2",
        False,
        12,
        humidity_files[3],
        temperature_files[3],
        5,
        10,
        20,
        6,
        True,
        True,
        "RepeatedKFold",
        6,
        60,
    )
    ml.main(
        steps,
        args.output_dir + "/ml/ml_kfold_2to2_3day",
        dataset_files_gain[4],
        "1To1",
        "2To2",
        False,
        12,
        humidity_files[4],
        temperature_files[4],
        5,
        10,
        20,
        6,
        True,
        True,
        "RepeatedKFold",
        6,
        60,
    )
    ml.main(
        steps,
        args.output_dir + "/ml/ml_kfold_2to2_2day",
        dataset_files_gain[5],
        "1To1",
        "2To2",
        False,
        12,
        humidity_files[5],
        temperature_files[5],
        5,
        10,
        20,
        6,
        True,
        True,
        "RepeatedKFold",
        6,
        60,
    )
    ml.main(
        steps,
        args.output_dir + "/ml/ml_kfold_2to2_1day",
        dataset_files_gain[6],
        "1To1",
        "2To2",
        False,
        12,
        humidity_files[6],
        temperature_files[6],
        5,
        10,
        20,
        6,
        True,
        True,
        "RepeatedKFold",
        6,
        60,
    )

    # ml.main(steps, args.output_dir + "/ml/ml_kfold_2to3_7day", dataset_files_gain[0], "1To1", "2To3", False, 12, humidity_files[0], temperature_files[0], 5, 10, 20, 6, True, True, "RepeatedKFold", 6, 60)
    # ml.main(steps, args.output_dir + "/ml/ml_kfold_2to3_6day", dataset_files_gain[1], "1To1", "2To3", False, 12, humidity_files[1], temperature_files[1], 5, 10, 20, 6, True, True, "RepeatedKFold", 6, 60)
    # ml.main(steps, args.output_dir + "/ml/ml_kfold_2to3_5day", dataset_files_gain[2], "1To1", "2To3", False, 12, humidity_files[2], temperature_files[2], 5, 10, 20, 6, True, True, "RepeatedKFold", 6, 60)
    # ml.main(steps, args.output_dir + "/ml/ml_kfold_2to3_4day", dataset_files_gain[3], "1To1", "2To3", False, 12, humidity_files[3], temperature_files[3], 5, 10, 20, 6, True, True, "RepeatedKFold", 6, 60)
    # ml.main(steps, args.output_dir + "/ml/ml_kfold_2to3_3day", dataset_files_gain[4], "1To1", "2To3", False, 12, humidity_files[4], temperature_files[4], 5, 10, 20, 6, True, True, "RepeatedKFold", 6, 60)
    # ml.main(steps, args.output_dir + "/ml/ml_kfold_2to3_2day", dataset_files_gain[5], "1To1", "2To3", False, 12, humidity_files[5], temperature_files[5], 5, 10, 20, 6, True, True, "RepeatedKFold", 6, 60)
    # ml.main(steps, args.output_dir + "/ml/ml_kfold_2to3_1day", dataset_files_gain[6], "1To1", "2To3", False, 12, humidity_files[6], temperature_files[6], 5, 10, 20, 6, True, True, "RepeatedKFold", 6, 60)

    # ml 1To1 1To2
    # ml.main(steps, args.output_dir + "/ml/ml_kfold_1to2_7day", dataset_files_gain[0], 1, 2, False, 12, humidity_files[0], temperature_files[0], 5, 10, 20, 6, True, True, "RepeatedKFold", 6, 60)
    # ml.main(steps, args.output_dir + "/ml/ml_kfold_1to2_6day", dataset_files_gain[1], 1, 2, False, 12, humidity_files[1], temperature_files[1], 5, 10, 20, 6, True, True, "RepeatedKFold", 6, 60)
    # ml.main(steps, args.output_dir + "/ml/ml_kfold_1to2_5day", dataset_files_gain[2], 1, 2, False, 12, humidity_files[2], temperature_files[2], 5, 10, 20, 6, True, True, "RepeatedKFold", 6, 60)
    # ml.main(steps, args.output_dir + "/ml/ml_kfold_1to2_4day", dataset_files_gain[3], 1, 2, False, 12, humidity_files[3], temperature_files[3], 5, 10, 20, 6, True, True, "RepeatedKFold", 6, 60)
    # ml.main(steps, args.output_dir + "/ml/ml_kfold_1to2_3day", dataset_files_gain[4], 1, 2, False, 12, humidity_files[4], temperature_files[4], 5, 10, 20, 6, True, True, "RepeatedKFold", 6, 60)
    # ml.main(steps, args.output_dir + "/ml/ml_kfold_1to2_2day", dataset_files_gain[5], 1, 2, False, 12, humidity_files[5], temperature_files[5], 5, 10, 20, 6, True, True, "RepeatedKFold", 6, 60)
    # ml.main(steps, args.output_dir + "/ml/ml_kfold_1to2_1day", dataset_files_gain[6], 1, 2, False, 12, humidity_files[6], temperature_files[6], 5, 10, 20, 6, True, True, "RepeatedKFold", 6, 60)

    # for window in [2, 30, 60, 60*12, 60*24]:
    #     ml.main(["QN", "ANSCOMBE", "LOG", "CENTER", "STFT", "STANDARDSCALER"], args.output_dir + "/ml/ml_kfold_2to2_7day_window_%d" % window, dataset_files[0], 1, 4, False, 1, None,
    #             None, 5, 10, 20, 6, True, True, "RepeatedKFold", None, window)
    #     ml.main(["QN", "ANSCOMBE", "LOG", "CENTER", "STFT", "STANDARDSCALER"], args.output_dir + "/ml/ml_kfold_2to2_6day_window_%d" % window, dataset_files[1], 1, 4, False, 1, None,
    #             None, 5, 10, 20, 6, True, True, "RepeatedKFold", None, window)
    #     ml.main(["QN", "ANSCOMBE", "LOG", "CENTER", "STFT", "STANDARDSCALER"], args.output_dir + "/ml/ml_kfold_2to2_5day_window_%d" % window, dataset_files[2], 1, 4, False, 1, None,
    #             None, 5, 10, 20, 6, True, True, "RepeatedKFold", None, window)
    #     ml.main(["QN", "ANSCOMBE", "LOG", "CENTER", "STFT", "STANDARDSCALER"], args.output_dir + "/ml/ml_kfold_2to2_4day_window_%d" % window, dataset_files[3], 1, 4, False, 1, None,
    #             None, 5, 10, 20, 6, True, True, "RepeatedKFold", None, window)
    #     ml.main(["QN", "ANSCOMBE", "LOG", "CENTER", "STFT", "STANDARDSCALER"], args.output_dir + "/ml/ml_kfold_2to2_3day_window_%d" % window, dataset_files[4], 1, 4, False, 1, None,
    #             None, 5, 10, 20, 6, True, True, "RepeatedKFold", None, window)
    #     ml.main(["QN", "ANSCOMBE", "LOG", "CENTER", "STFT", "STANDARDSCALER"], args.output_dir + "/ml/ml_kfold_2to2_2day_window_%d" % window, dataset_files[5], 1, 4, False, 1, None,
    #             None, 5, 10, 20, 6, True, True, "RepeatedKFold", None, window)
    #     ml.main(["QN", "ANSCOMBE", "LOG", "CENTER", "STFT", "STANDARDSCALER"], args.output_dir + "/ml/ml_kfold_2to2_1day_window_%d" % window, dataset_files[6], 1, 4, False, 1, None,
    #             None, 5, 10, 20, 6, True, True, "RepeatedKFold", None, window)
    #
    # for f0 in [1, 2, 3, 4, 5, 6, 7, 8, 9, 10]:
    #     f0_str = str(f0).replace(".", "_")
    #     ml.main(["QN", "ANSCOMBE", "LOG", "CENTER", "CWT", "STANDARDSCALER"], args.output_dir + "/ml/ml_kfold_2to2_7day_wf0_%s" % f0_str, dataset_files[0], 1, 4, False, 1, None,
    #             None, 5, 10, 20, 6, True, True, "RepeatedKFold", f0, 256)
    #     ml.main(["QN", "ANSCOMBE", "LOG", "CENTER", "CWT", "STANDARDSCALER"], args.output_dir + "/ml/ml_kfold_2to2_6day_wf0_%s" % f0_str, dataset_files[1], 1, 4, False, 1, None,
    #             None, 5, 10, 20, 6, True, True, "RepeatedKFold", f0, 256)
    #     ml.main(["QN", "ANSCOMBE", "LOG", "CENTER", "CWT", "STANDARDSCALER"], args.output_dir + "/ml/ml_kfold_2to2_5day_wf0_%s" % f0_str, dataset_files[2], 1, 4, False, 1, None,
    #             None, 5, 10, 20, 6, True, True, "RepeatedKFold", f0, 256)
    #     ml.main(["QN", "ANSCOMBE", "LOG", "CENTER", "CWT", "STANDARDSCALER"], args.output_dir + "/ml/ml_kfold_2to2_4day_wf0_%s" % f0_str, dataset_files[3], 1, 4, False, 1, None,
    #             None, 5, 10, 20, 6, True, True, "RepeatedKFold", f0, 256)
    #     ml.main(["QN", "ANSCOMBE", "LOG", "CENTER", "CWT", "STANDARDSCALER"], args.output_dir + "/ml/ml_kfold_2to2_3day_wf0_%s" % f0_str, dataset_files[4], 1, 4, False, 1, None,
    #             None, 5, 10, 20, 6, True, True, "RepeatedKFold", f0, 256)
    #     ml.main(["QN", "ANSCOMBE", "LOG", "CENTER", "CWT", "STANDARDSCALER"], args.output_dir + "/ml/ml_kfold_2to2_2day_wf0_%s" % f0_str, dataset_files[5], 1, 4, False, 1, None,
    #             None, 5, 10, 20, 6, True, True, "RepeatedKFold", f0, 256)
    #     ml.main(["QN", "ANSCOMBE", "LOG", "CENTER", "CWT", "STANDARDSCALER"], args.output_dir + "/ml/ml_kfold_2to2_1day_wf0_%s" % f0_str, dataset_files[6], 1, 4, False, 1, None,
    #             None, 5, 10, 20, 6, True, True, "RepeatedKFold", f0, 256)

    # ml.main(args.output_dir + "/ml/ml_l2out_2to2_7day", dataset_files[0], 1, 4, False, 1, None, None, 5, 10, 20, 6, True, True, "LeaveTwoOut")
    # ml.main(args.output_dir + "/ml/ml_l2out_2to2_6day", dataset_files[1], 1, 4, False, 1, None, None, 5, 10, 20, 6, True, True, "LeaveTwoOut")
    # ml.main(args.output_dir + "/ml/ml_l2out_2to2_5day", dataset_files[2], 1, 4, False, 1, None, None, 5, 10, 20, 6, True, True, "LeaveTwoOut")
    # ml.main(args.output_dir + "/ml/ml_l2out_2to2_4day", dataset_files[3], 1, 4, False, 1, None, None, 5, 10, 20, 6, True, True, "LeaveTwoOut")
    # ml.main(args.output_dir + "/ml/ml_l2out_2to2_3day", dataset_files[4], 1, 4, False, 1, None, None, 5, 10, 20, 6, True, True, "LeaveTwoOut")
    # ml.main(args.output_dir + "/ml/ml_l2out_2to2_2day", dataset_files[5], 1, 4, False, 1, None, None, 5, 10, 20, 6, True, True, "LeaveTwoOut")
    # ml.main(args.output_dir + "/ml/ml_l2out_2to2_1day", dataset_files[6], 1, 4, False, 1, None, None, 5, 10, 20, 6, True, True, "LeaveTwoOut")
    #
    # ml.main(args.output_dir + "/ml/ml_l1out_2to2_7day", dataset_files[0], 1, 4, False, 1, None, None, 5, 10, 20, 6, True, True, "LeaveOneOut")
    # ml.main(args.output_dir + "/ml/ml_l1out_2to2_6day", dataset_files[1], 1, 4, False, 1, None, None, 5, 10, 20, 6, True, True, "LeaveOneOut")
    # ml.main(args.output_dir + "/ml/ml_l1out_2to2_5day", dataset_files[2], 1, 4, False, 1, None, None, 5, 10, 20, 6, True, True, "LeaveOneOut")
    # ml.main(args.output_dir + "/ml/ml_l1out_2to2_4day", dataset_files[3], 1, 4, False, 1, None, None, 5, 10, 20, 6, True, True, "LeaveOneOut")
    # ml.main(args.output_dir + "/ml/ml_l1out_2to2_3day", dataset_files[4], 1, 4, False, 1, None, None, 5, 10, 20, 6, True, True, "LeaveOneOut")
    # ml.main(args.output_dir + "/ml/ml_l1out_2to2_2day", dataset_files[5], 1, 4, False, 1, None, None, 5, 10, 20, 6, True, True, "LeaveOneOut")
    # ml.main(args.output_dir + "/ml/ml_l1out_2to2_1day", dataset_files[6], 1, 4, False, 1, None, None, 5, 10, 20, 6, True, True, "LeaveOneOut")

    # ml 1To1 1To2
    # ml.main(args.output_dir + "/ml/ml_kfold_1to2_7day", dataset_files[0], 1, 2, False, 1, None, None, 5, 10, 20, 6, True, True, "RepeatedKFold")
    # ml.main(args.output_dir + "/ml/ml_kfold_1to2_6day", dataset_files[1], 1, 2, False, 1, None, None, 5, 10, 20, 6, True, True, "RepeatedKFold")
    # ml.main(args.output_dir + "/ml/ml_kfold_1to2_5day", dataset_files[2], 1, 2, False, 1, None, None, 5, 10, 20, 6, True, True, "RepeatedKFold")
    # ml.main(args.output_dir + "/ml/ml_kfold_1to2_4day", dataset_files[3], 1, 2, False, 1, None, None, 5, 10, 20, 6, True, True, "RepeatedKFold")
    # ml.main(args.output_dir + "/ml/ml_kfold_1to2_3day", dataset_files[4], 1, 2, False, 1, None, None, 5, 10, 20, 6, True, True, "RepeatedKFold")
    # ml.main(args.output_dir + "/ml/ml_kfold_1to2_2day", dataset_files[5], 1, 2, False, 1, None, None, 5, 10, 20, 6, True, True, "RepeatedKFold")
    # ml.main(args.output_dir + "/ml/ml_kfold_1to2_1day", dataset_files[6], 1, 2, False, 1, None, None, 5, 10, 20, 6, True, True, "RepeatedKFold")

    # ml.main(args.output_dir + "/ml/ml_l2out_1to2_7day", dataset_files[0], 1, 2, False, 1, None, None, 5, 10, 20, 6, True, True, "LeaveTwoOut")
    # ml.main(args.output_dir + "/ml/ml_l2out_1to2_6day", dataset_files[1], 1, 2, False, 1, None, None, 5, 10, 20, 6, True, True, "LeaveTwoOut")
    # ml.main(args.output_dir + "/ml/ml_l2out_1to2_5day", dataset_files[2], 1, 2, False, 1, None, None, 5, 10, 20, 6, True, True, "LeaveTwoOut")
    # ml.main(args.output_dir + "/ml/ml_l2out_1to2_4day", dataset_files[3], 1, 2, False, 1, None, None, 5, 10, 20, 6, True, True, "LeaveTwoOut")
    # ml.main(args.output_dir + "/ml/ml_l2out_1to2_3day", dataset_files[4], 1, 2, False, 1, None, None, 5, 10, 20, 6, True, True, "LeaveTwoOut")
    # ml.main(args.output_dir + "/ml/ml_l2out_1to2_2day", dataset_files[5], 1, 2, False, 1, None, None, 5, 10, 20, 6, True, True, "LeaveTwoOut")
    # ml.main(args.output_dir + "/ml/ml_l2out_1to2_1day", dataset_files[6], 1, 2, False, 1, None, None, 5, 10, 20, 6, True, True, "LeaveTwoOut")
    #
    # ml.main(args.output_dir + "/ml/ml_l1out_1to2_7day", dataset_files[0], 1, 2, False, 1, None, None, 5, 10, 20, 6, True, True, "LeaveOneOut")
    # ml.main(args.output_dir + "/ml/ml_l1out_1to2_6day", dataset_files[1], 1, 2, False, 1, None, None, 5, 10, 20, 6, True, True, "LeaveOneOut")
    # ml.main(args.output_dir + "/ml/ml_l1out_1to2_5day", dataset_files[2], 1, 2, False, 1, None, None, 5, 10, 20, 6, True, True, "LeaveOneOut")
    # ml.main(args.output_dir + "/ml/ml_l1out_1to2_4day", dataset_files[3], 1, 2, False, 1, None, None, 5, 10, 20, 6, True, True, "LeaveOneOut")
    # ml.main(args.output_dir + "/ml/ml_l1out_1to2_3day", dataset_files[4], 1, 2, False, 1, None, None, 5, 10, 20, 6, True, True, "LeaveOneOut")
    # ml.main(args.output_dir + "/ml/ml_l1out_1to2_2day", dataset_files[5], 1, 2, False, 1, None, None, 5, 10, 20, 6, True, True, "LeaveOneOut")
    # ml.main(args.output_dir + "/ml/ml_l1out_1to2_1day", dataset_files[6], 1, 2, False, 1, None, None, 5, 10, 20, 6, True, True, "LeaveOneOut")