| hclust
HCLUST
hierarchical clustering, faster version
[LABELS, DENDRO] = HCLUST(D,TYPE,K,OLD)
DENDRO = HCLUST(D,TYPE)
| Input | | D | dissimilarity matrix | | TYPE | string name of clustering criterion (optional) 's' or 'single' : single linkage (default) 'c' or 'complete' : complete linkage 'a' or 'average' : average linkage (weighted over cluster sizes) | | K | number of clusters (optional) | | OLD | Logical, if TRUE return dendrogram in PRTools format. Default | | FALSE | |
| Output | | LABELS | vector with labels | | DENDRO | arrsy with dendrogram |
Description Computation of cluster labels and a clustering dendrogram for the objects with a given dissimilarity matrix D. K is the desired number of | clusters. | The dendrogram may be plotted by PRTools's PLOTDG or by | | Matlab's | DENDROGRAM.
DENDRO = HCLUST(D,TYPE)
As in this case no clustering level is supplied, just the entire dendrogram is returned. The first row now contains the object indices.
Faster and more advanced tools for cluster analysis may be found in the ClusterTools toolbox. Example(s)
a = gendats([25 25],20,5); % 50 points in 20-dimensional feature space
d = sqrt(distm(a)); % Euclidean distances
dendg = hclustf(d,'complete'); % dendrogram
plotdg(dendg)
lab = hclust(d,'complete',2); % labels
confmat(lab,getlabels(a)); % confusion matrix
See also
plotdg, prkmeans, kcentres, modeseek, emclust, dendrogram, | | This file has been automatically generated. If badly readable, use the help-command in Matlab. |
|