Re: [R] Which cluster function can be used to cluster a correlaiton matrix?

From: Martin Maechler <maechler_at_stat.math.ethz.ch>
Date: Wed 28 Dec 2005 - 21:59:01 EST

>>>>> "Vincent" == Vincent Deng <bioflash@gmail.com>
>>>>> on Wed, 28 Dec 2005 17:34:27 +0800 writes:

    Vincent> Hi, I'd got a matrix of correaltion values. Which     Vincent> cluster method can I use to cluster it?

almost everyone.

The clue is to transform correlations to dissimilarities. There are several choices for that, and it depends on the context what you should do.
If 'Cx' is your correlation matrix, reasonable possibilities are

  Dx <- as.dist(1 - Cx)
  Dx <- as.dist(1 - abs(Cx))
  Dx <- as.dist(sqrt(1 - Cx^2))

and then use hclust(), agnes(), pam(), [the latter two from package 'cluster'], ...
with 'Dx' as dissimilarity



R-help@stat.math.ethz.ch mailing list
https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html Received on Wed Dec 28 22:05:26 2005

This archive was generated by hypermail 2.1.8 : Fri 03 Mar 2006 - 03:41:42 EST