mvc (version 1.3)

mvcsph: Multi-View Clustering using Spherical k-Means for categorical data.

Description

Multi-View Clustering using Spherical k-Means for categorical data. See: S. Bickel, T. Scheffer: Multi-View Clustering, ICDM 04. Hierachical clustering used to determine the initial centers for k-Means

Usage

mvcsph(view1, view2, k=Inf, startView="view1", nthresh=20, doOutput=F, doDebug=F, plotFile="Rplots.pdf")

Arguments

view1

View number one, a data frame with the same row names as view2. All columns numeric. Entries are natural numbers, starting from 1.

view2

View number two, a data frame with the same row names as view1. All columns numeric. Entries are natural numbers, starting from 1.

k

The maximum number of clusters to create

startView

The view on which to perform the initial E step, one of "view1", "view2"

nthresh

The number of iterations to run without improvement of the objective function

doOutput

Whether output to the console should be done

doDebug

Whether debug output to the console should be done (implies normal output)

plotFile

File name where the hierarchical clustering plot should be stored

Value

Examples

# Demo program, showing how to run Multi-
# View Clustering using Spherical k-Means 
# AM, 2011

# load toy data 'toyView1' and 'toyView2'
data(toyViews)

mvcsph(
view1=toyView1,
view2=toyView2,
nthresh=20,
k=4,
startView="view1"
)

}