spe.tree.AdaptiveLinearSmoother

class spe.tree.AdaptiveLinearSmoother

ABC class all adaptive linear smoother models should inherit from.

Enforces subclasses to have the method necessary to be valid model inputs to cp_adaptive_smoother(), cp_bagged() when full_refit is True.

Methods

__init__(*args, **kwargs)
get_linear_smoother(X)	Get fitted adaptive linear smoother matrix \(S(W)\).
predict(X[, tr_idx, ts_idx, y_refit])	Compute \(S(W)y_{refit}\) as predictions.

__init__(*args, **kwargs)

abstract get_linear_smoother(X)

Get fitted adaptive linear smoother matrix \(S(W)\).

predict(X, tr_idx=None, ts_idx=None, y_refit=None)

Compute \(S(W)y_{refit}\) as predictions.

Computes predictions as adaptive linear smoothing where \(S(W)\) is the output of instance’s AdaptiveLinearSmoother().