skglm.datafits.WeightedQuadratic#

class skglm.datafits.WeightedQuadratic(sample_weights)[source]#

Weighted Quadratic datafit to handle sample weights.

The datafit reads:

`1 / (2 xx \sum_(i=1)^(n_"samples") weights_i) \sum_(i=1)^(n_"samples") weights_i (y_i - (Xw)_i)^ 2`
Attributes:
Xtwyarray, shape (n_features,)

Pre-computed quantity used during the gradient evaluation. Equal to X.T @ (samples_weights * y).

sample_weightsarray, shape (n_samples,)

Weights for each sample.

__init__(sample_weights)[source]#

Methods

__init__(sample_weights)

full_grad_sparse(X_data, X_indptr, ...)

get_global_lipschitz(X, y)

get_global_lipschitz_sparse(X_data, ...)

get_lipschitz(X, y)

get_lipschitz_sparse(X_data, X_indptr, ...)

get_spec()

Specify the numba types of the class attributes.

gradient(X, y, Xw)

gradient_scalar(X, y, w, Xw, j)

gradient_scalar_sparse(X_data, X_indptr, ...)

initialize(X, y)

Pre-computations before fitting on X and y.

initialize_sparse(X_data, X_indptr, X_indices, y)

Pre-computations before fitting on X and y when X is a sparse matrix.

intercept_update_step(y, Xw)

inverse_link(Xw)

Inverse link function (identity by default).

params_to_dict()

Get the parameters to initialize an instance of the class.

raw_grad(y, Xw)

raw_hessian(y, Xw)

value(y, w, Xw)

Value of datafit at vector w.