fastcan

Fast canonical correlation analysis based search algorithm.

Classes

FastCan([n_features_to_select, ...])	Forward feature selector according to the sum of squared canonical correlation coefficients (SSC).
LazyFastCan([n_features_to_select, ...])	Lazy version of FastCan selector.

Functions

minibatch(X, y[, n_features_to_select, ...])	Feature selection using fastcan.FastCan with mini batches.
refine(selector[, drop, max_iter, verbose])	Two-stage refining for the results of fastcan.FastCan.