CLUSTHC

Classifier for hierarchical clustering

     W = CLUSTHC(A,TYPE)
     W = A*CLUSTHC(TYPE);
     LAB = B*W*LABELD

Input
A	Dataset, typically labeld with cluster labels obtained by a  hierarchical clustering of the given TYPE, see CLUSTH.
TYPE	Linkage rule: 's', 'a', 'r' or 'c'. Alternatively  'single','average','central' or 'complete'. Default 'single'.
B	New objects (dataset or doubles) to be assigned to the clusters  in A.

Output
LAB	Estimated labels of B.
W	Classifier(s) assigning new objects B using the linkage rule  defined by TYPE.

Description

This routine is useful to find a classifier for assigning new objects to  clusters found by a hierarchical clustering.

It is assumed that A is a dataset of which the labels refer to clusters  found by a hierarchical cluster procedure CLUSTH with linkage type TYPE.  The classifier W assigns new objects B to the most neighboring cluster  according to the cluster distance defined by TYPE.

Example(s)

 x = +gendatclust(1000);
 scattern(x);
 lab = x*clusth(10);
 w = prdataset(x,lab)*clusthc;
 plotc(w);

See also

datasets, mappings, labeld, clusth, labeld,