Description
Optimal Partitioning of Similarity Relations.
Description
Contains a set of algorithms for creating partitions and coverings of objects largely based on operations on (dis)similarity relations (or matrices). There are several iterative re-assignment algorithms optimizing different goodness-of-clustering criteria. In addition, there are covering algorithms 'clique' which derives maximal cliques, and 'maxpact' which creates a covering of maximally compact sets. Graphical analyses and conversion routines are also included.