check_clone_preserves_forecaster_params¶

def check_clone_preserves_forecaster_params(forecaster) -> None:
    """Check sklearn's clone() preserves init parameters.

    Delegates the shallow-param and nested-estimator contract to the shared,
    family-agnostic ``check_clone_preserves_params`` (which uses ``_safe_equal``
    and so tolerates DataFrame/array-valued params whose ``==`` is not a bool).
    Adds the forecaster-specific check for ``(name, estimator, columns)``
    3-tuples (e.g. ``ColumnForecaster``): the trailing ``columns`` element must
    survive clone unchanged.

    Parameters
    ----------
    forecaster : BaseForecaster
        Forecaster instance

    Raises
    ------
    AssertionError
        If cloned forecaster has different parameters

    """
    check_clone_preserves_params(forecaster)

    forecaster_clone = clone(forecaster)
    original_params = forecaster.get_params(deep=False)
    cloned_params = forecaster_clone.get_params(deep=False)

    for key, orig_val in original_params.items():
        cloned_val = cloned_params[key]
        if not (isinstance(orig_val, list) and orig_val and isinstance(orig_val[0], tuple)):
            continue
        for i, (orig_item, cloned_item) in enumerate(zip(orig_val, cloned_val, strict=True)):
            if not (isinstance(orig_item, tuple) and isinstance(cloned_item, tuple)):
                continue
            if len(orig_item) == 3 and len(cloned_item) == 3:
                orig_cols, cloned_cols = orig_item[-1], cloned_item[-1]
                assert _safe_equal(orig_cols, cloned_cols), (
                    f"Parameter {key}[{i}] columns: {cloned_cols!r} != {orig_cols!r}"
                )

    assert forecaster_clone is not forecaster, "clone() should create new instance"