sparseprcomp: Fast PCA using the irlba library for SVD

In zachmayer/r2vec: Text data to numeric vectors

Description

This functions performs principle components analysis via SVD, using the irlba library. Note that it is not possible to center and scale a sparse matrix. Folding in the eigenvalues might help compensate for this.

Usage

sparseprcomp(x, n = 5, retx = FALSE, fold_in_eigens = FALSE, ...)

Arguments

x	a sparse matrix
n	number of principle components to calculate
retx	whether to return the principle components of the input data
fold_in_eigens	whether to add the eigenvectors to the principle component rotations
...	arguments passed to irlbda

Value

a sparse.prcomp object, which inherits from prcomp

zachmayer/r2vec documentation built on May 4, 2019, 9:05 p.m.