This is a cluster method based on DICM We need at least two inputs: a term-document matrix and the clusters number. As a example, you can use the a.pny as input file to clusters the documents.