---

class sts_select.mrmr.MRMRBase(scorer, n_features=30, preselected_features=None)

Applies the mRMR scoring algorithm given a BaseScorer. Inherits the BaseEstimator class from sklearn.

Parameters:

• scorer (BaseScorer) – A BaseScorer object.

• n_features – The number of features to select.

• preselected_features – A list of features to preselect.

distance_X_y(original)

Averages X-y score pairs for a given feature. :param original: :return:

fit(X, y)

Fits the mRMR algorithm to the data given the scoring algorithm. Adapted from fast-mRMR without optimizations (Ramírez-Gallego et al.)

Parameters:

• X – The data to fit (not used).

• y – The labels to fit (not used).

score_label(vec, label)

Returns the score for each feature and the overall label. :param vec: :param label: :return:

score_lfs_candidate(X, candidates, lastFeatureSelected)

Returns the score between candidates and the last feature selected. :param X: :param candidates: :param lastFeatureSelected: :return:

transform(X)