utils.py

# Authors:
#       Helion du Mas des Bourboux <helion.dumasdesbourboux'at'thalesgroup.com>
#       Thomas Courtat <thomas.courtat'at'thalesgroup.com>
#
# MIT License
#
# Copyright (c) 2022 THALES
#   All Rights Reserved.
#
# Permission is hereby granted, free of charge, to any person obtaining a copy
# of this software and associated documentation files (the "Software"), to deal
# in the Software without restriction, including without limitation the rights
# to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
# copies of the Software, and to permit persons to whom the Software is
# furnished to do so, subject to the following conditions:
#
# The above copyright notice and this permission notice shall be included in
# all copies or substantial portions of the Software.
#
# THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
# IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
# FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
# AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
# LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
# OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
# SOFTWARE.
#
# 2022 November 25

import time
from tensorflow.keras.callbacks import Callback
from numpy.random import choice


class TimeHistory(Callback):
    """Keras callback tto monitor execution time during training"""

    def on_train_begin(self, logs={}):
        self.times = []

    def on_epoch_begin(self, epoch, logs={}):
        self.epoch_time_start = time.time()

    def on_epoch_end(self, epoch, logs={}):
        self.times.append(time.time() - self.epoch_time_start)


def split_dataset(
    data,
    labels,
    p_train=0.8,
    p_valid=0.0,
    p_test=0.0,
):
    """Function to split and shuffle a dataset into train, validation and test
    splits.

    Arguments:
    data (tensor): input dataset
    labels (tensor): label vector associated to data
    p_train (float): proportion of training data
    p_valid (float): proportion of validation data
    p_test (float): proportion of test data
    """

    def aux(D_, L_, p_, idx_):

        n_tot_examples_ = D_.shape[0]
        n_taken_examples_ = int(n_tot_examples_ * p_)
        chosen_idx_ = choice(idx_, size=n_taken_examples_, replace=False)
        free_indexes_ = list(set(idx_) - set(chosen_idx_))

        return D_[chosen_idx_], L_[chosen_idx_], chosen_idx_, free_indexes_

    D = data
    L = labels

    idx = range(0, D.shape[0])
    X_train, Y_train, train_idx, free_idx = aux(D, L, p_train, idx)

    idx = free_idx
    X_valid, Y_valid, valid_idx, free_idx = aux(D, L, p_valid, idx)

    idx = free_idx
    X_test, Y_test, test_idx, free_idx = aux(D, L, p_test, idx)

    return (
        X_train,
        Y_train,
        train_idx,
        X_valid,
        Y_valid,
        valid_idx,
        X_test,
        Y_test,
        test_idx,
    )