EMA (version 1.4.3)

clustering.kmeans: Kmeans and hierarchical clustering

Description

Kmeans clustering to summarize the genes information and hierarchical clustering on the kmeans' groups

Usage

clustering.kmeans(data, N = 100, iter.max = 20,
title = "Kmeans - Hierarchical Clustering",
dist.s = "pearson", dist.g = "pearsonabs", method = "ward")

Arguments

data

Expression matrix, genes on rows and samples on columns

N

The number of a priori clusters for the kmeans

iter.max

The maximum number of iterations allowed for the kmeans clustering

title

The plot title

dist.s

The distance used for the sample clustering

dist.g

The distance used for the genes clustering

method

The linkage used for both clusterings

Value

• A list with the kmeans object and the two hierarchical clusterings.
• c.km
  An object of class 'kmeans'.
• c.sample
  An object of class 'agnes'. The hierarchical clustering on samples
• c.kcenters
  An object of class 'agnes'. The hierarchical clustering on the kmeans centers

Details

The goal of this analysis is to first summarizes the genes information using the kmeans clustering. Then, a two-ways clustering is performed using the center of each kmean groups, and all the samples.

See Also

kmeans,agnes

Examples

data(marty)
##Example on 100 genes for 5 classes
clustering.kmeans(marty[1:100,], N=5)