I've written a new package for identifying clusters of observations by
means of the unsupervised learning K-means algorithm. For details see here:
The new package --named "kmeans" -- is currently not available at the
gretl package server as it depends on a new built-in function which will
not be be available before the next gretl release.
However, those of you who would like to use it right now, simply need to
do the following:
1. Get the latest gretl snapshot, and
2. obtain the package file (incl. the documentation and sample script)
Please report any remaining bugs on the mailing list.