GeneralizedLinearRegression.score

GeneralizedLinearRegression.score(X, y, coords=None, **kwargs)

Score Bayesian \(R^2\) for Gaussian models; skip for other families.

Parameters:

• X (xr.DataArray) – Input features.

• y (xr.DataArray) – Observed targets to score against the posterior predictive mean.

• coords (dict[str, Any] | None) – Coordinate names for named dimensions. Forwarded to PyMCModel.score().

• **kwargs – Reserved for subclass extensions.

Returns:

Per-unit Bayesian \(R^2\) scores for Gaussian models; None for non-identity families.

Return type:

pd.Series or None