stableclus {FactoClass}R Documentation

Stable Clusters from K-means

Description

It performs stable clusters from several partitions from K-means changing the initial points. It uses the coordinates of an previous principal axes method

Usage

stableclus(dudi,part=2,k.clust=2,ff.clus=NULL,bplot=TRUE,kmns=FALSE)

Arguments

dudi object of type dudi
part the number of partitions to obtain
k.clust the number of clusters each partition
ff.clus the number of ended clusters, if is NULL (default), it asks in the console the values ff.clus
bplot if is TRUE, the barplot of the stable clusters is ploted
kmns when hmns=TRUE, a kmeansW is performed with argument centers equal to the centroids of the stable classes retained

Details

Lebart, L. and Morineau, A. and Piron, M. (1995) Statisitique exploratoire multidimensionnelle, Paris.

Value

list with:

class vector indicating the cluster of each element
stacl vector containing the stable clusters in descending order

Author(s)

Carlos Andres Arias caariasr@unal.edu.co>, Campo Elias Pardo cepardot@unal.edu.co, Diana Carolina Zarate dczarated@unal.edu.co

Examples

 data(BreedsDogs)
 acm <- dudi.acm(BreedsDogs[,-7],scannf = FALSE,nf=10)
 stableclus(acm,2,3,4)

[Package FactoClass version 1.0.1 Index]