_cnn.py

"""Time Convolutional Neural Network (CNN) for regression."""

__author__ = ["AurumnPegasus", "achieveordie", "hadifawaz1999"]
__all__ = ["CNNRegressor"]

import gc
import os
import time
from copy import deepcopy

from sklearn.utils import check_random_state

from aeon.networks import CNNNetwork
from aeon.regression.deep_learning.base import BaseDeepRegressor
from aeon.utils.validation._dependencies import _check_soft_dependencies


class CNNRegressor(BaseDeepRegressor):
    """Time Series Convolutional Neural Network (CNN).

    Adapted from the implementation used in [1]_.

    Parameters
    ----------
    n_layers        : int, default = 2,
        the number of convolution layers in the network
    kernel_size    : int or list of int, default = 7,
        kernel size of convolution layers, if not a list, the same kernel size
        is used for all layer, len(list) should be n_layers
    n_filters       : int or list of int, default = [6, 12],
        number of filters for each convolution layer, if not a list, the same n_filters
        is used in all layers.
    avg_pool_size   : int or list of int, default = 3,
        the size of the average pooling layer, if not a list, the same
        max pooling size is used
        for all convolution layer
    output_activation   : str, default = "linear",
        the output activation for the regressor
    activation      : str or list of str, default = "sigmoid",
        keras activation function used in the model for each layer,
        if not a list, the same
        activation is used for all layers
    padding         : str or list of str, default = 'valid',
        the method of padding in convolution layers, if not a list,
        the same padding used
        for all convolution layers
    strides         : int or list of int, default = 1,
        the strides of kernels in the convolution and max pooling layers,
        if not a list, the same strides are used for all layers
    dilation_rate   : int or list of int, default = 1,
        the dilation rate of the convolution layers, if not a list,
        the same dilation rate is used all over the network
    use_bias        : bool or list of bool, default = True,
        condition on whether or not to use bias values for convolution layers,
        if not a list, the same condition is used for all layers
    random_state    : int, default = 0
        seed to any needed random actions
    n_epochs       : int, default = 2000
        the number of epochs to train the model
    batch_size      : int, default = 16
        the number of samples per gradient update.
    verbose         : boolean, default = False
        whether to output extra information
    loss            : string, default="mean_squared_error"
        fit parameter for the keras model
    optimizer       : keras.optimizer, default=keras.optimizers.Adam(),
    metrics         : list of strings, default=["accuracy"],
    callbacks       : keras.callbacks, default=model_checkpoint to save best
                      model on training loss
    file_path       : file_path for the best model (if checkpoint is used as callback)
    save_best_model     : bool, default = False
        Whether or not to save the best model, if the
        modelcheckpoint callback is used by default,
        this condition, if True, will prevent the
        automatic deletion of the best saved model from
        file and the user can choose the file name
    save_last_model     : bool, default = False
        Whether or not to save the last model, last
        epoch trained, using the base class method
        save_last_model_to_file
    best_file_name      : str, default = "best_model"
        The name of the file of the best model, if
        save_best_model is set to False, this parameter
        is discarded
    last_file_name      : str, default = "last_model"
        The name of the file of the last model, if
        save_last_model is set to False, this parameter
        is discarded

    Notes
    -----
    Adapted from the implementation from Fawaz et. al
    https://github.com/hfawaz/dl-4-tsc/blob/master/classifiers/cnn.py

    References
    ----------
    .. [1] Zhao et. al, Convolutional neural networks for time series classification,
    Journal of Systems Engineering and Electronics, 28(1):2017.

    Examples
    --------
    >>> from aeon.regression.deep_learning import CNNRegressor
    >>> from aeon.datasets import make_example_3d_numpy
    >>> X, y = make_example_3d_numpy(n_cases=10, n_channels=1, n_timepoints=12,
    ...                              return_y=True, regression_target=True,
    ...                              random_state=0)
    >>> rgs = CNNRegressor(n_epochs=20, bacth_size=4) # doctest: +SKIP
    >>> rgs.fit(X, y) # doctest: +SKIP
    CNNRegressor(...)
    """

    def __init__(
        self,
        n_layers=2,
        kernel_size=7,
        n_filters=None,
        avg_pool_size=3,
        activation="sigmoid",
        padding="valid",
        strides=1,
        dilation_rate=1,
        n_epochs=2000,
        batch_size=16,
        callbacks=None,
        file_path="./",
        save_best_model=False,
        save_last_model=False,
        best_file_name="best_model",
        last_file_name="last_model",
        verbose=False,
        loss="mse",
        output_activation="linear",
        metrics=None,
        random_state=None,
        use_bias=True,
        optimizer=None,
    ):
        _check_soft_dependencies("tensorflow")
        super(CNNRegressor, self).__init__(
            batch_size=batch_size,
        )
        self.n_layers = n_layers
        self.avg_pool_size = avg_pool_size
        self.padding = padding
        self.n_filters = n_filters
        self.kernel_size = kernel_size
        self.file_path = file_path
        self.save_best_model = save_best_model
        self.save_last_model = save_last_model
        self.best_file_name = best_file_name
        self.last_file_name = last_file_name
        self.strides = strides
        self.dilation_rate = dilation_rate
        self.callbacks = callbacks
        self.n_epochs = n_epochs
        self.batch_size = batch_size
        self.verbose = verbose
        self.loss = loss
        self.output_activation = output_activation
        self.metrics = metrics
        self.random_state = random_state
        self.activation = activation
        self.use_bias = use_bias
        self.optimizer = optimizer
        self.history = None
        self._network = CNNNetwork(
            n_layers=self.n_layers,
            kernel_size=self.kernel_size,
            n_filters=self.n_filters,
            avg_pool_size=self.avg_pool_size,
            activation=self.activation,
            padding=self.padding,
            strides=self.strides,
            dilation_rate=self.dilation_rate,
            use_bias=self.use_bias,
            random_state=self.random_state,
        )

    def build_model(self, input_shape, **kwargs):
        """Construct a compiled, un-trained, keras model that is ready for training.

        In aeon, time series are stored in numpy arrays of shape (d,m), where d
        is the number of dimensions, m is the series length. Keras/tensorflow assume
        data is in shape (m,d). This method also assumes (m,d). Transpose should
        happen in fit.

        Parameters
        ----------
        input_shape : tuple
            The shape of the data fed into the input layer, should be (m,d)

        Returns
        -------
        output : a compiled Keras Model
        """
        import tensorflow as tf
        from tensorflow import keras

        tf.random.set_seed(self.random_state)

        input_layer, output_layer = self._network.build_network(input_shape, **kwargs)

        output_layer = keras.layers.Dense(units=1, activation=self.output_activation)(
            output_layer
        )

        self.optimizer_ = (
            keras.optimizers.Adam() if self.optimizer is None else self.optimizer
        )

        model = keras.models.Model(inputs=input_layer, outputs=output_layer)

        model.compile(
            loss=self.loss,
            optimizer=self.optimizer_,
            metrics=self.metrics,
        )
        return model

    def _fit(self, X, y):
        """Fit the regressor on the training set (X, y).

        Parameters
        ----------
        X : np.ndarray of shape = (n_instances (n), n_channels (d), series_length (m))
            The training input samples.
        y : np.ndarray of shape n
            The training data target values.

        Returns
        -------
        self : object
        """
        import tensorflow as tf

        # Transpose to conform to Keras input style.
        X = X.transpose(0, 2, 1)

        check_random_state(self.random_state)

        self.input_shape = X.shape[1:]
        self.training_model_ = self.build_model(self.input_shape)

        if self.verbose:
            self.training_model_.summary()

        self.file_name_ = (
            self.best_file_name if self.save_best_model else str(time.time_ns())
        )

        self.callbacks_ = (
            [
                tf.keras.callbacks.ModelCheckpoint(
                    filepath=self.file_path + self.file_name_ + ".hdf5",
                    monitor="loss",
                    save_best_only=True,
                ),
            ]
            if self.callbacks is None
            else self.callbacks
        )

        self.history = self.training_model_.fit(
            X,
            y,
            batch_size=self.batch_size,
            epochs=self.n_epochs,
            verbose=self.verbose,
            callbacks=self.callbacks_,
        )

        try:
            self.model_ = tf.keras.models.load_model(
                self.file_path + self.file_name_ + ".hdf5", compile=False
            )
            if not self.save_best_model:
                os.remove(self.file_path + self.file_name_ + ".hdf5")
        except FileNotFoundError:
            self.model_ = deepcopy(self.training_model_)

        if self.save_last_model:
            self.save_last_model_to_file(file_path=self.file_path)

        gc.collect()
        return self

    @classmethod
    def get_test_params(cls, parameter_set="default"):
        """Return testing parameter settings for the estimator.

        Parameters
        ----------
        parameter_set : str, default="default"
            Name of the set of test parameters to return, for use in tests. If no
            special parameters are defined for a value, will return `"default"` set.
            For regressors, a "default" set of parameters should be provided for
            general testing, and a "results_comparison" set for comparing against
            previously recorded results if the general set does not produce suitable
            probabilities to compare against.

        Returns
        -------
        params : dict or list of dict, default={}
            Parameters to create testing instances of the class.
            Each dict are parameters to construct an "interesting" test instance, i.e.,
            `MyClass(**params)` or `MyClass(**params[i])` creates a valid test instance.
            `create_test_instance` uses the first (or only) dictionary in `params`.
        """
        param = {
            "n_epochs": 10,
            "batch_size": 4,
            "avg_pool_size": 4,
        }

        return [param]