nnpiv.diagnostics.relative_wellposedness_effective_sieve_diagnostic
- nnpiv.diagnostics.relative_wellposedness_effective_sieve_diagnostic(A, C, C_prime, e_g, *, sieve_grid=None, eta_grid=None, enforce_nested_rff=True, **kwargs)[source]
Run
kappa_effover a sieve and eta path.- Parameters
A (array-like) – Inputs passed to
relative_wellposedness_effective_diagnostic(). Arrays must share the same number of rows.C (array-like) – Inputs passed to
relative_wellposedness_effective_diagnostic(). Arrays must share the same number of rows.C_prime (array-like) – Inputs passed to
relative_wellposedness_effective_diagnostic(). Arrays must share the same number of rows.e_g (array-like of shape (n,) or (n, 1)) – Estimated first-stage error direction.
sieve_grid (iterable, optional) – Sieve values interpreted as in
relative_wellposedness_sieve_diagnostic().eta_grid (float or iterable of float, optional) – Stabilization values for path evaluation.
enforce_nested_rff (bool, default=True) – Reuse nested RFF prefixes when possible.
**kwargs – Additional keyword arguments forwarded to
relative_wellposedness_effective_diagnostic().
- Returns
Dictionary with
rows(path-level outputs) andsummary(aggregated metadata and maxima).- Return type
dict
Notes
This is the post-estimation analogue of
relative_wellposedness_sieve_diagnostic().