Skip to content

validate_splitter_data

yohou.utils.validate_data.validate_splitter_data(splitter, y, X_actual) 

validate_splitter_data(
    splitter: BaseSplitter,
    y: pl.DataFrame,
    X_actual: pl.DataFrame | None,
) -> tuple[pl.DataFrame, pl.DataFrame | None]

validate_splitter_data(
    splitter: BaseSplitter,
    y: None,
    X_actual: pl.DataFrame | None,
) -> tuple[None, pl.DataFrame | None]

Validate and prepare input data for time series splitters.

Checks that y and X_actual have valid "time" columns, consistent panel structure, and matching panel groups. When y is provided the time interval is inferred and stored on the splitter as interval_.

Parameters

Name Type Description Default
splitter BaseSplitter

The splitter instance. interval_ will be set on it when y is not None.

 required
y DataFrame or None

Target time series. When None, only X_actual is validated.

 required
X_actual DataFrame or None

Exogenous features. When None, only y is validated.

 required

Returns

Type Description
tuple of (pl.DataFrame or None, pl.DataFrame or None)

Validated (y, X_actual) pair.

Raises

Type Description
ValueError

If time columns are missing, panel groups are inconsistent, or input validation fails.

See Also

Source Code

View on GitHub

Show/Hide source 
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
def validate_splitter_data(
    splitter: BaseSplitter, y: pl.DataFrame | None, X_actual: pl.DataFrame | None
) -> tuple[pl.DataFrame | None, pl.DataFrame | None]:
    """Validate and prepare input data for time series splitters.

    Checks that ``y`` and ``X_actual`` have valid ``"time"`` columns, consistent
    panel structure, and matching panel groups.  When ``y`` is provided the
    time interval is inferred and stored on the splitter as ``interval_``.

    Parameters
    ----------
    splitter : BaseSplitter
        The splitter instance.  ``interval_`` will be set on it when
        ``y`` is not ``None``.
    y : pl.DataFrame or None
        Target time series.  When ``None``, only ``X_actual`` is validated.
    X_actual : pl.DataFrame or None
        Exogenous features.  When ``None``, only ``y`` is validated.

    Returns
    -------
    tuple of (pl.DataFrame or None, pl.DataFrame or None)
        Validated ``(y, X_actual)`` pair.

    Raises
    ------
    ValueError
        If time columns are missing, panel groups are inconsistent, or
        input validation fails.

    See Also
    --------
    - [`BaseSplitter`][yohou.model_selection.split.BaseSplitter] : Base class for time series CV splitters.
    - [`check_inputs`][yohou.utils.validation.check_inputs] : Low-level input validation helper.

    """
    if y is not None:
        check_time_column(y)
        check_panel_internal_consistency(y, "y")

    if X_actual is not None:
        check_time_column(X_actual)
        check_panel_internal_consistency(X_actual, "X_actual")

        if y is not None:
            check_panel_groups_match(y, X_actual)

    # Type narrowing: check_inputs requires non-None y
    if y is not None:
        interval = check_inputs(y, X_actual)
        splitter.interval_ = interval

    return y, X_actual