strategycat: strategycat function

In Clustericat: This package performs clustering for categorical dataset.

Description

This function allows to create a new strategycat object. This function needs a data matrix as mandatory argument.

Usage

  strategycat(data, nb_init = 5,
    stop_criterion = 20 * ncol(data),
    partition = partitioncat(data))

Arguments

data	Input data as matrix of non-zero integers.
nb_init	Integer of the number of times where a MCMC chain is trating. By default 5 MCMC chains are initializated.
stop_criterion	Integer corresponding to the number of successive iterations of the MCMC chain where if no best model are fund then the algorithm is stopped. By default it takes the values of 20x d.
partition	Vector of integer. The element j takes the values of the block where the variable j is affected for the initialization of the MCMC chains for all the classes. By default the function partitioncat is called.

Value

Return an object of strategycat class.

Examples

# Simple example with binary data i
data("dentist")

#creation of a strategycat object with the constructor by default
defaultstrategycat <- strategycat(dentist)

#presentation of the adjustment parameters of the strategycat object
summary(defaultstrategycat)

#creation of new object strategycat
newstrategycat <- strategycat(dentist, nb_init=100, stop_criterion=10, partition=c(1,2,3,1,2))

#presentation of the new object
summary(newstrategycat)

Clustericat documentation built on May 2, 2019, 5:45 p.m.