mvcmb: Multi-View Clustering using mixture of categoricals EM.
In mvc: Multi-View Clustering

Description Usage Arguments Value Examples

Multi-View Clustering using mixture of categoricals EM. See: S. Bickel, T. Scheffer: Multi-View Clustering, ICDM 04.

1 2	mvcmb(view1, view2, k=Inf, startView="view1", nthresh=20, doOutput=F, doDebug=F)

`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`	String designating 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)

A list reporting the final clustering, with names finalIndices, agreementRate, indicesSv, indicesOv. They designate final cluster indices as a vector, as well as agreement rate of the two views, and the individual indices given by the two views over the course of iterations as data frames.

{
# Demo program, showing how to run Multi-
# View Clustering using Mixture of Binomials EM.
# AM, 2011

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

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

}