opt.clust.num: Find optimal number of clusters.

In rezakj/scSeqR: scSeqR; a toolkit to analyze single cell sequencing data types (i.e scRNA-seq) and large numeric matrix files.

Find optimal number of clusters.

Description

This function takes an object of class scSeqR and finds optimal number of clusters based on three methods.

Usage

opt.clust.num(x = NULL, max.clust = 20, gap.stat.nboot = 100,
  verbose = TRUE, clust.type = "tsne", opt.method = "silhouette")

Arguments

x	An object of class scSeqR.
max.clust	Maximum number of clusters allowed to optimize, defult = 20.
gap.stat.nboot	Number of boots, defult = 100 (500 is desired but slower).
verbose	If TRUE will show the processing.
clust.type	Choose from "tsne","pca" or "distance", defult = "tsne".
opt.method	Choose from "elbow.wss", "silhouette", "gap.stat" or "", defult = "all".

Value

An object of class scSeqR.

Examples

## Not run: 
opt.clust.num(my.obj, max.clust = 10, clust.type = "tsne", opt.method = "silhouette")

## End(Not run)

rezakj/scSeqR documentation built on March 28, 2022, 12:17 p.m.