svd: Approximate or randomized singular value decomposition

In jenzopr/singlecellutils: Utility functions for single-cell data

Description

Approximate SVD is performed using the implementation from irlba, randomized SVD is performed using the implementation from rsvd.

Usage

randomized_svd(object, exprs_values, n_dims, features = NULL,
  skip = NULL, seed = NULL, ...)

approximate_svd(object, exprs_values, n_dims, features = NULL,
  skip = NULL, seed = NULL, ...)

Arguments

object	A SingleCellExperiment object.
exprs_values	String indicating which assay contains the data that should be used to perform SVD.
n_dims	The number of approximate singular values to calculate.
features	A character vector (of feature names), a logical vector or numeric vector (of indices) specifying the features to use for SVD. The default of NULL will use all features.
skip	A numeric vector indicating which singular values to set to zero (and remove).
seed	A numeric seed to initialize the random number generator.
...	Additional arguments passed on to rsvd or irlba.

Value

A matrix of dimension ncol(object) x dims(object) - length(skip).

jenzopr/singlecellutils documentation built on June 12, 2019, 2:51 a.m.