nCluster: nCluster

In VCCRI/CIDR: Clustering through Imputation and Dimensionality Reduction

Description

outputs the Calinski-Harabasz Index versus number of clusters plot.

Usage

## S4 method for signature 'scData'
nCluster(object, n = NULL, nPC = NULL,
  cMethod = "ward.D2")

Arguments

object	an scData class object.
n	maximum number of clusters; if NULL (default), it is set to be nPC*2+2.
nPC	the number of PCs used in clustering; by default 4.
cMethod	hierarchical clustering method; by default "ward.D2".

Details

outputs the Calinski-Harabasz Index versus number of clusters plot, useful for deciding the number of clusters. Since CIDR automatically decides the number of clusters, this method is optional and only used if the user wants to alter the default number of clusters.

Examples

example(cidr)

VCCRI/CIDR documentation built on May 9, 2019, 9:41 p.m.