Re: clustering in NCL

From: Saji Hameed <saji.nh_at_nyahnyahspammersnyahnyah>
Date: Fri Nov 04 2011 - 05:07:30 MDT

Hi Axel,

If you want to try out self organizing maps for clustering, you can find my
ruby bindings at https://github.com/sajinh/SOMPAK_4R
The input is a text-formatted file consisting of several rows of the
vectors of interest (time series or collapsed spatial maps)

The ruby bindings in turn are based on the C-implementation. It should be
possible to write a Fortran Wrapper for the C-bindings and call it from NCL.

saji

On Fri, Nov 4, 2011 at 7:26 PM, Seifert Axel <Axel.Seifert@dwd.de> wrote:

>
> Hi NCL folks,
>
> about a year ago there was a short discussion on the same subject in
> ncl-talk. Any new experience since then that somebody would like to share?
>
> I would especially be interested in a density-based spatial clustering
> method like DBSCAN
>
> http://en.wikipedia.org/wiki/DBSCAN
>
> If anybody has or knows a free FORTRAN code that I could plug into NCL, I
> would also very much appreciate the hint.
>
> Best, Axel
>
> _______________________________________________
> ncl-talk mailing list
> List instructions, subscriber options, unsubscribe:
> http://mailman.ucar.edu/mailman/listinfo/ncl-talk
>

_______________________________________________
ncl-talk mailing list
List instructions, subscriber options, unsubscribe:
http://mailman.ucar.edu/mailman/listinfo/ncl-talk
Received on Fri Nov 4 05:07:39 2011

This archive was generated by hypermail 2.1.8 : Fri Nov 04 2011 - 08:43:41 MDT