nnpiv.diagnostics.relative_wellposedness_from_data

nnpiv.diagnostics.relative_wellposedness_from_data(data, *, A, B=None, C, C_prime, mask_s=None, mask_t=None, **kwargs)

Run the pre-estimation diagnostic from dataset-level selectors.

Parameters: data : DataFrame, mapping, or 2D array-like

Source container holding all blocks.

A, C, C_prime

Block selectors for each argument required by relative_wellposedness_diagnostic().

B

Optional selector accepted only for interface consistency with (A, B, C, C') notation. It is intentionally ignored by this diagnostic.

Supported selector forms: - callable selector(data) -> array-like - direct array-like matrix/vector - DataFrame/mapping columns (string or list of strings) - DataFrame/array-style integer selectors (int, list[int], slice)

mask_s, mask_tarray-like, str, or callable, optional

Optional subset selectors for stage-specific sample restrictions.

**kwargs

Forwarded to relative_wellposedness_diagnostic().

Returns

Output dictionary returned by relative_wellposedness_diagnostic().

Return type

dict

Notes

A, C, and C_prime must resolve to arrays with the same row count.