nnls.project: Get for mapping weights (H) for a set of samples to a latent...

In zdebruine/scNMF: Fast Non-negative matrix factorization toolkit for single cell data

Description

This function will minimize mean((samples - W.model

Usage

nnls.project(
  samples,
  W.model,
  n.threads = 0,
  block.size = NULL,
  verbose = TRUE,
  scd.tol = 1e-08,
  scd.max.iter = 500
)

Arguments

samples	Samples to be projected onto W.model
W.model	NMF factor model (genes x factors) for mapping to cells
n.threads	Number of threads/CPUs to use. Default to 0 (all cores)
block.size	How many samples to project concurrently (default is NULL)
verbose	Boolean
scd.tol	Tolerance of the sequential coordinate NNLS fit (default 1e-8)
scd.max.iter	Maximum iterations of sequential coordinate descent within the NNLS fit (default 500)

Value

A matrix of mapping weights (H). The returned matrix will be coerced to sparse format if the input was sparse.

zdebruine/scNMF documentation built on Jan. 1, 2021, 1:50 p.m.