heterog.parti: heterogeneity of a partition obtained with the method hclust.geo.H.

Description

This function calculate the heterogeneity of a partition as the sum of the heterogeneity of all classes obtained withheterog.Ck.

Usage

1
heterog.parti(dist.var, dist.geo, partition, wt, alpha)

Arguments

dist.var

a square symetric matrix with the distances between individuals measured on the variables.

dist.geo

a square symetric matrix with the geographical distances between individuals.

partition

a vector of size n giving the belonging of each individuals to each class of individuals.

wt

a vector of size n whose values indicate weights of individuals on quantitative variables

alpha

a real between 0 and 1 to fit the global criterion to optimize

Details

Note that in the calcul of Inertia, the two matrices of distances are normalized (divided by their own maximum).

Value

res a list containing the following results : the Inertia of the partition based on dist.var, the Inertia of the partition based on dist.geo, the heterogeneity of the partition as defined in description.


Questions? Problems? Suggestions? or email at ian@mutexlabs.com.

All documentation is copyright its authors; we didn't write any of that.