Skip to content

check_fit_sets_attributes

yohou.testing.transformer.check_fit_sets_attributes(transformer, X, y=None)

Check fit() sets required attributes.

Validates that fit() creates feature_names_in_, n_features_in_, and _observation_horizon attributes as expected by sklearn conventions.

Parameters

Name Type Description Default
transformer BaseTransformer

Unfitted transformer instance

 required
X DataFrame

Training data with "time" column

 required
y DataFrame

Target data for supervised transformers

 None

Raises

Type Description
AssertionError

If required attributes are not set after fit()

Source Code

View on GitHub

Show/Hide source 
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
def check_fit_sets_attributes(transformer, X: pl.DataFrame, y: pl.DataFrame | None = None) -> None:
    """Check fit() sets required attributes.

    Validates that fit() creates feature_names_in_, n_features_in_,
    and _observation_horizon attributes as expected by sklearn conventions.

    Parameters
    ----------
    transformer : BaseTransformer
        Unfitted transformer instance
    X : pl.DataFrame
        Training data with "time" column
    y : pl.DataFrame, optional
        Target data for supervised transformers

    Raises
    ------
    AssertionError
        If required attributes are not set after fit()

    """
    transformer_clone = clone(transformer)
    transformer_clone.fit(X, y)

    # Check sklearn-required attributes
    assert hasattr(transformer_clone, "feature_names_in_"), "fit() must set feature_names_in_ attribute"
    assert hasattr(transformer_clone, "n_features_in_"), "fit() must set n_features_in_ attribute"

    # Check yohou-required attributes
    assert hasattr(transformer_clone, "_observation_horizon"), "fit() must set _observation_horizon attribute"

    # Validate values
    expected_features = [col for col in X.columns if col != "time"]
    assert list(transformer_clone.feature_names_in_) == expected_features, (
        f"feature_names_in_ mismatch: {transformer_clone.feature_names_in_} vs {expected_features}"
    )

    assert transformer_clone.n_features_in_ == len(expected_features), (
        f"n_features_in_ should be {len(expected_features)}, got {transformer_clone.n_features_in_}"
    )