Python clustering code
Dan Ensign
1/18/2007

Please note that some documentation in the .py files may be a bit outdated.

clusterers are found in 'Algorithms.py'
TestObject is found in 'Algorithms.py'

a contact map example is found in cmaps2.py. This and TestObject should provide
examples for how to make your own objects that can be used with KMeansKlusterer.

A simple script that sorts TestObjects is: runCluster.py.


To do?
Save the data *before* starting to cluster?

Or cluster, then select new data to test on the generators?

Eliminate parallel support in the clustering algorithms? Instead, we could write our
data objects to do their comparisons, etc., in a parallel fashion. For something like 
contact maps this may increase the overhead, but it might not be that bad. The code
would be simpler.