Aluode/PerceptionLabPortable

1# Authors: The MNE-Python contributors.2# License: BSD-3-Clause3# Copyright the MNE-Python contributors.4 5try:6    from sklearn.utils.validation import validate_data7except ImportError:8    from sklearn.utils.validation import check_array, check_X_y9 10    # Use a limited version pulled from sklearn 1.711    def validate_data(12        _estimator,13        /,14        X="no_validation",15        y="no_validation",16        reset=True,17        validate_separately=False,18        skip_check_array=False,19        **check_params,20    ):21        """Validate input data and set or check feature names and counts of the input.22 23        This helper function should be used in an estimator that requires input24        validation. This mutates the estimator and sets the `n_features_in_` and25        `feature_names_in_` attributes if `reset=True`.26 27        .. versionadded:: 1.628 29        Parameters30        ----------31        _estimator : estimator instance32            The estimator to validate the input for.33 34        X : {array-like, sparse matrix, dataframe} of shape \35                (n_samples, n_features), default='no validation'36            The input samples.37            If `'no_validation'`, no validation is performed on `X`. This is38            useful for meta-estimator which can delegate input validation to39            their underlying estimator(s). In that case `y` must be passed and40            the only accepted `check_params` are `multi_output` and41            `y_numeric`.42 43        y : array-like of shape (n_samples,), default='no_validation'44            The targets.45 46            - If `None`, :func:`~sklearn.utils.check_array` is called on `X`. If47            the estimator's `requires_y` tag is True, then an error will be raised.48            - If `'no_validation'`, :func:`~sklearn.utils.check_array` is called49            on `X` and the estimator's `requires_y` tag is ignored. This is a default50            placeholder and is never meant to be explicitly set. In that case `X` must51            be passed.52            - Otherwise, only `y` with `_check_y` or both `X` and `y` are checked with53            either :func:`~sklearn.utils.check_array` or54            :func:`~sklearn.utils.check_X_y` depending on `validate_separately`.55 56        reset : bool, default=True57            Whether to reset the `n_features_in_` attribute.58            If False, the input will be checked for consistency with data59            provided when reset was last True.60 61            .. note::62 63            It is recommended to call `reset=True` in `fit` and in the first64            call to `partial_fit`. All other methods that validate `X`65            should set `reset=False`.66 67        validate_separately : False or tuple of dicts, default=False68            Only used if `y` is not `None`.69            If `False`, call :func:`~sklearn.utils.check_X_y`. Else, it must be a tuple70            of kwargs to be used for calling :func:`~sklearn.utils.check_array` on `X`71            and `y` respectively.72 73            `estimator=self` is automatically added to these dicts to generate74            more informative error message in case of invalid input data.75 76        skip_check_array : bool, default=False77            If `True`, `X` and `y` are unchanged and only `feature_names_in_` and78            `n_features_in_` are checked. Otherwise, :func:`~sklearn.utils.check_array`79            is called on `X` and `y`.80 81        **check_params : kwargs82            Parameters passed to :func:`~sklearn.utils.check_array` or83            :func:`~sklearn.utils.check_X_y`. Ignored if validate_separately84            is not False.85 86            `estimator=self` is automatically added to these params to generate87            more informative error message in case of invalid input data.88 89        Returns90        -------91        out : {ndarray, sparse matrix} or tuple of these92            The validated input. A tuple is returned if both `X` and `y` are93            validated.94        """95        no_val_X = isinstance(X, str) and X == "no_validation"96        no_val_y = y is None or (isinstance(y, str) and y == "no_validation")97 98        if no_val_X and no_val_y:99            raise ValueError("Validation should be done on X, y or both.")100 101        default_check_params = {"estimator": _estimator}102        check_params = {**default_check_params, **check_params}103 104        if skip_check_array:105            if not no_val_X and no_val_y:106                out = X107            elif no_val_X and not no_val_y:108                out = y109            else:110                out = X, y111        elif not no_val_X and no_val_y:112            out = check_array(X, input_name="X", **check_params)113        elif no_val_X and not no_val_y:114            out = check_array(y, input_name="y", **check_params)115        else:116            if validate_separately:117                # We need this because some estimators validate X and y118                # separately, and in general, separately calling check_array()119                # on X and y isn't equivalent to just calling check_X_y()120                # :(121                check_X_params, check_y_params = validate_separately122                if "estimator" not in check_X_params:123                    check_X_params = {**default_check_params, **check_X_params}124                X = check_array(X, input_name="X", **check_X_params)125                if "estimator" not in check_y_params:126                    check_y_params = {**default_check_params, **check_y_params}127                y = check_array(y, input_name="y", **check_y_params)128            else:129                X, y = check_X_y(X, y, **check_params)130            out = X, y131 132        return out133 134 135def _check_n_features_3d(estimator, X, reset):136    """Set the `n_features_in_` attribute, or check against it on an estimator.137 138    Sklearn takes n_features from X.shape[1], but we need X.shape[-1]139 140    Parameters141    ----------142    estimator : estimator instance143        The estimator to validate the input for.144 145    X : {ndarray, sparse matrix} of shape ([n_epochs], n_samples, n_features)146        The input samples.147 148    reset : bool149        If True, the `n_features_in_` attribute is set to `X.shape[1]`.150        If False and the attribute exists, then check that it is equal to151        `X.shape[1]`. If False and the attribute does *not* exist, then152        the check is skipped.153        .. note::154        It is recommended to call reset=True in `fit` and in the first155        call to `partial_fit`. All other methods that validate `X`156        should set `reset=False`.157    """158    n_features = X.shape[-1]159    if reset:160        estimator.n_features_in_ = n_features161        return162 163    if not hasattr(estimator, "n_features_in_"):164        # Skip this check if the expected number of expected input features165        # was not recorded by calling fit first. This is typically the case166        # for stateless transformers.167        return168 169    if n_features != estimator.n_features_in_:170        raise ValueError(171            f"X has {n_features} features, but {estimator.__class__.__name__} "172            f"is expecting {estimator.n_features_in_} features as input."173        )174