clusterRMSD: Compute the RMSD per cluster

In LTLA/bluster: Clustering Algorithms for Bioconductor

Compute the RMSD per cluster

Description

Compute the root mean-squared deviation (RMSD) for each cluster.

Usage

clusterRMSD(x, clusters, sum = FALSE)

Arguments

x	Numeric matrix containing observations in rows and variables in columns.
clusters	Vector containing the assigned cluster for each observation.
sum	Logical scalar indicating whether to compute the sum of squares.

Details

The RMSD for each cluster is a measure of its dispersion; clusters with large internal heterogeneity will have high RMSDs and are good candidates for further subclustering.

Value

Numeric vector of RMSD values per cluster. If sum=TRUE, a numeric vector of the sum of squares per cluster is returned instead.

Author(s)

Aaron Lun

Examples

x <- matrix(rnorm(10000), ncol=10)
kout <- kmeans(x, 5)
clusterRMSD(x, kout$cluster)

LTLA/bluster documentation built on Aug. 20, 2023, 5:39 a.m.