Feature selection

Wrappers for the feature selectors of scikit-matter.

class equisolve.numpy.feature_selection.CUR(recompute_every=1, k=1, tolerance=1e-12, n_to_select=None, score_threshold=None, score_threshold_type='absolute', progress_bar=False, full=False, random_state=0)

Bases: GreedySelector

Transformer that performs Greedy Feature Selection with CUR.

Refer to skmatter.feature_selection.CUR for full documentation.

class equisolve.numpy.feature_selection.FPS(initialize=0, n_to_select=None, score_threshold=None, score_threshold_type='absolute', progress_bar=False, full=False, random_state=0)

Bases: GreedySelector

Transformer that performs Greedy Feature Selection using Farthest Point Sampling.

Refer to skmatter.feature_selection.FPS for full documentation.