[BioC] Kmeans help

[Carlos J. Gil Bellosta]

Hello,

Instead of k-means you could use pam in the cluster package. According to

http://stat.ethz.ch/R-manual/R-patched/library/cluster/html/pam.html

you can use the diss parameter to indicate pam that you are dealing
with a distance (dissimilary)  matrix instead of the original data
frame.

Best regards,

Carlos J. Gil Bellosta
http://www.datanalytics.com


2010/3/23 sidahmed BENABDERRAHMANE <sidahmed.benabderrahmane at loria.fr>:
> Dear all,
> Let me ask my question on this list because I am a new user of R.
> I have a problem with the Kmeans clustering method . In fact, having a
> symmetric matrix:
>
>
>        g1      g2      .. ...       gn
> g1     1        0,3                   0,6
> g2      0,3      1                    0,4
> .         ......................................
> .         ....................................
> ..     ...........................................
> gn    0,6       0,4                       1
>
> I should like to apply Kmeans on this matrix, knowing that there is no
> relation  R(object, attribute) but R(object, object) only.
> Does the Kmeans algorithm works in this case?
> Could you give me a syntax for testingKmeans on this matrix.
>
> Thanks
>
> ----
>
> --
> ------------------------------------------------------------------
> Sidahmed BENABDERRAHMANE
> PhD, LORIA  France
>
> _______________________________________________
> Bioconductor mailing list
> Bioconductor at stat.math.ethz.ch
> https://stat.ethz.ch/mailman/listinfo/bioconductor
> Search the archives:
> http://news.gmane.org/gmane.science.biology.informatics.conductor
>