Ridge

class superglm.Ridge(

lambda1: float | Literal['auto'] | None = None,

features: str | list[str] | None = None,

)

Bases: object

Ridge penalty: lambda1 * ||beta||_2^2 / 2.

Parameters:

lambda1float, {“auto”}, or None

Regularisation strength. None disables shrinkage and "auto" explicitly requests automatic calibration.

prox_group(

bg: NDArray,

group: GroupSlice,

step: float,

) → NDArray

Closed-form proximal operator for a single group.

prox(

beta: NDArray,

groups: list[GroupSlice],

step: float,

) → NDArray

Closed-form proximal operator: beta / (1 + step * lambda1).

eval(

beta: NDArray,

groups: list[GroupSlice],

) → float

Penalty value: lambda1 * ||beta||_2^2 / 2.