Two remarks:

1) On cores/consensus: Similar methods tend to produce similar results,
   whatever the data say. I think these core/consensus methods work well
   only if the compared CA methods are sufficiently distinct, and then
   there is a good chance that no "cores" are found (if the clustering in
   the data is not obvious).

2) On clustering with R1=R2=R3=R. k-means clustering implicitly assumes
   clusters to have unit matrix correlation. So transforming the data to
   unit covariance and then applying 3-means will give clusters with
   approximately R1=R2=R3=R. May be even better with a Gausiian mixture
   model where covariance matrices of the clusters are restricted to cI,
   where I is unit matrix and c may depend on the cluster. This again has
   to be applied to data which is sphered, i.e. transformed to unit
   covariance first. I hope this "covariance model" can be found in mclust,
   mentioned previously in this discussion.

Christian Hennig



--
***********************************************************************
Christian Hennig
Seminar fuer Statistik, ETH-Zentrum (LEO), CH-8092 Zuerich (current)
and Fachbereich Mathematik-SPST/ZMS, Universitaet Hamburg
[log in to unmask], http://stat.ethz.ch/~hennig/
[log in to unmask], http://www.math.uni-hamburg.de/home/hennig/
#######################################################################
ich empfehle www.boag.de