FRtest: Friedman-Rafsky (FR) test

In JCVenterInstitute/FRmatch: Cell type matching in single-cell RNA-sequencing data using FR-Match

Friedman-Rafsky (FR) test

Description

FR test is a multivariate generalization of nonparametric two-sample test. This function is an implementation with customized options, including a visualization of the minimum spanning tree (MST).

Usage

FRtest(
  samp1,
  samp2,
  use.cosine = FALSE,
  binary = FALSE,
  binary.cutoff = 2,
  plot.MST = FALSE,
  col = c("#F0E442", "#56B4E9"),
  label.names = c("Sample 1", "Sample 2"),
  vertex.size = 5,
  edge.width = 1,
  ...
)

Arguments

samp1	Numeric matrix or data frame for Sample 1. Rows are multivariate dimensions, and columns are samples. E.g. gene by cell.
samp2	Numeric matrix or data frame for Sample 2.
use.cosine	An option if to use cosine distance. Logical variable. By default (FALSE), Euclidean distance is used.
binary	An option if to use binary values. Logical variable. Default: FALSE. If TRUE, use binary.cutoff to dichotomize samp1 and samp2.
binary.cutoff	Numeric value for binary cutoff. Binary value = 1 if greater than binary.cutoff, 0 otherwise. Default: 2.
plot.MST	Boolean variable indicating if to plot the minimum spanning tree (MST). Default: FALSE.
col	Character vector of length two for customized colors of the nodes in MST. Default: c("#F0E442", "#56B4E9").
label.names	Character vector of length two for customized names of the two samples. Default: c("Sample 1","Sample 2").
vertex.size, edge.width, ...	Additional plotting parammeters passed to plot.igraph. Default: vertex.size=5, edge.width=1.

Value

Test statistics and p-values.

runs	Total number of subtrees.
runs.samp1	Number of subtrees of Sample 1.
runs.samp2	Number of subtrees of Sample 2.
stat	The standardized FR statistic.
p.value	P-value of the FR test.

Examples

## Not run: 
samp1 <- matrix(rnorm(100),nrow=5)
samp2 <- matrix(rnorm(100),nrow=5)
FRtest(samp1, samp2)
FRtest(samp1, samp2, use.cosine=TRUE)
FRtest(samp1, samp2, plot.MST=TRUE, main="Minimum spanning tree plot")
FRtest(samp1, samp2, binary=TRUE, binary.cutoff=1)

## End(Not run)

JCVenterInstitute/FRmatch documentation built on Dec. 15, 2022, 2:30 p.m.