predict#

SuperGLM.predict(
X: object,
offset: NDArray | None = None,
*,
random_effects: str = 'conditional',
) → NDArray#

Predict the response mean for new data.

Parameters:
Xpandas or eager Polars DataFrame

Eager input features with the same columns used during fitting.

offsetNDArray or None

Optional offset added to the linear predictor before applying the inverse link.

random_effects{“conditional”, “population”}

Whether to include fitted random-effect and factor-smooth deviations. Population prediction sets all such contributions to zero.

Returns:
NDArray

Predicted mean on the response scale (inverse-link of eta).