cluster: Clustering of multinomial data

In bomarkussen/clusterMultinomial: Hierarchical clustering of multinomial data

Description

Agglomerative clustering of observations from a collection of multinomial variables.

Usage

cluster(Y_data,X_data=NULL,logit.model=NULL,ordering=NULL,method="pvalue",verbose=FALSE)

Arguments

Y_data	data.frame with responses, which all should be factors
X_data	data.frame with possible predictors
logit.model	formula for logistic regression separating the observations
ordering	numeric used for cluster ordering
method	If pvalue, then p-values are used for selection. Otherwise the likelihood ratio is used.
verbose	output of algorithm tracking

Details

To be added

Value

Object of class multinomClust, which is a LIST with 19 elements. Among these

Y_data	Data frame with response data
X_data	Data frame with covariate data
logit.model	Model formula for logistic model
category.names	Names of variables in Y_data
clusters.sequence	List of list of clusters
p.list	Vector with p-values for agglomerative clustering
number.clusters	Integer with estimated number of clusters

Author(s)

Bo Markussen ([e-mail])

Examples

## Not run: 
data(SES)
SEScluster <- cluster(SES[,-(1:5)],SES[,1:5],.~GL,SES$dpi)

## End(Not run)

# Plot
data(SEScluster)
plot(SEScluster)

bomarkussen/clusterMultinomial documentation built on May 4, 2019, 6:39 p.m.