reduceDimensions: Reduce dimension using Principal Components Analysis via...

In JEFworks-Lab/veloviz: VeloViz: RNA-velocity informed 2D embeddings for visualizing cell state trajectories

Reduce dimension using Principal Components Analysis via svds from RSpectra

Description

Reduce dimension using Principal Components Analysis via svds from RSpectra

Usage

reduceDimensions(
  matnorm,
  center = TRUE,
  scale = TRUE,
  max.ods.genes = 2000,
  nPCs = 50,
  verbose = TRUE,
  plot = FALSE,
  details = FALSE
)

Arguments

matnorm	matrix on which to perform PCA
center	logical to mean center gene expression before PCA, default = TRUE
scale	logical to scale gene expression variance before PCA, default = TRUE
max.ods.genes	number of most highly expressed overdispersed genes to include, default = 2000
nPCs	number of principal components to reduce to return, default = 50
verbose	logical for verbosity setting, default = TRUE
plot	plot singular values vs number of components
details	logical to return pca object, default = FALSE

Value

matrix of cell scores in nPCs components

Examples

data(vel)
curr <- vel$current

curr.norm <- normalizeDepth(curr)
curr.norm <- log10(curr.norm+1)
reduceDimensions(curr.norm, nPCs=3)

JEFworks-Lab/veloviz documentation built on Sept. 14, 2022, 4:03 p.m.