qha (version 0.0.8)

tableclass: Correspondence Analysis Factorial Planes

Description

It suggests the number of axes to using in a cluster analysis

Usage

tableclass(tabla,numaxes=2,numclass=2)

Arguments

tabla

object of type data frame or matrix with coordinates of factorial analysis

numaxes

number of dimensions for the classification. Default 2

numclass

number of class. Default 2

Value

Details

To understand the differences between elements of differents class, the planes factoriales are insufficient. It is necessary, therefore, to proceed to the classification of the individuals in the vectorial space of the first factors of the analysis. To decide the number of axes to using, Barbary (1996) proposed a procedure which consists of realizing the classification several times changing the number of axes and using as criterion the number of individuals of the class with major quantity of individuals. The analysis to interpreting is that one that originates the minor number of individuals in the majority class. The function tableclass realizes the classification several times y it proposes a number of axes to the classification.

Examples

data(ratingTV)

# Fuzzy Correspondence Analysis 
fuzzyb<-fuzzy(datos=ratingTV$tab, vectorc=c(15,15,15,15,15,15))  
20

tabla_clases<-tableclass(fuzzyb$li, numaxes=4, numclass=7)
tabla_clases