ClusterEval {svcR}R Documentation

Computation of evaluation

Description

Set of functions making evaluation. It requires a class column in data.

Usage

## S4 method for signature 'findModelCluster':
ClusterEval.Pvalue(x=NULL)

## S4 method for signature 'matrix':
DataSampling(DatMat=matrix())

## S4 method for signature 'findModelCluster':
ClusterEval.crossval(x=NULL)

## S4 method for signature 'findModelCluster':
ClusterEval.final(x=NULL)

## S4 method for signature 'findModelCluster':
ClusterEval.scalable(x=NULL)

## S4 method for signature 'findModelCluster':
ClusterEval.clusterable(x=NULL)

Arguments

x a findModelCluster object
DatMat iris matrix obtain from a run with findModelCluster function

Details

DataSampling, ClusterEval.crossval, ClusterEval.final, ClusterEval.scalable and ClusterEval.clusterable are based only on iris data. ClusterEval.scalable make evaluation building a matrix adding line by line to current matrix from iris matrix.ClusterEval.clusterable makes evaluation with a set of q paramter values and a set of nu parameter values.

Value

return a numeric or a list of numeric meaning a precision value between 0 and 100 for the method ClusterEval.scalable and ClusterEval.clusterable it return a list of precision values.

Author(s)

Nicolas Turenne - INRA France nicolas.turenne@jouy.inra.fr

References

N.Turenne , Some Heuristics to speed-up Support Vector Clustering , technical report 2006, INRA, France http://migale.jouy.inra.fr/~turenne/svc.pdf

Examples


## exemple with iris data

# generation of the svc model
fmc = findModelCluster.Test(); 

# catch evaluation value
P = ClusterEval.Pvalue( fmc );


[Package svcR version 1.4.2 Index]