[S] summary of replies re weighted multidimensional scaling

Rob Tibshirani (tibs@stat.stanford.edu)
Wed, 2 Dec 1998 08:34:47 -0800 (PST)


I rec'd many useful messages re my request for a weighted multidimensional
scaling routine.

Doug Martin ,Doug Clarkson, Brian Ripley, Andreas Buja, Forrest Young
and Tim Johnson all replied.

Here is a summary:

-Brian's isoMDS in the Splus library MASS does not take weights, but
could be modified by incorporating wts in the monotone regression.
One would have to prove that this solves the wted problem.

-Forrest Young has a FORTRAN routine ASCAL on his website
http://forrest.psych.unc.edu

-Tim Johnson has an Splus routine for wted MDS on his website
http://www.psych.uiuc.edu/~tjohnson/tjohnson.html
based on De Leeuw's (1977) SMACOF. I have already tried it and it seems
to work well!

-Andreas Buja pointed out his new MDS system xgvi.
The completed paper on the subject is in one of

xgvis.ps, xgvis.ps.gz, xgvis.pdf, xgvis.pdf.gz

from
ftp://ftp.research.att.com/dist/xgobi/

-Finally, Jan De Leeuw has a superb paper

Fitting distances by least squares

on his website
http://www.stat.ucla.edu/~deleeuw/work/research.phtml

Thank you very much to everyone who helped out.

Rob Tibshirani
**********************************************
Rob Tibshirani, Dept of Health Research & Policy
and Dept of Statistics
HRP Redwood Bldg
Stanford University
Stanford, California 94305-5405

phone: HRP: 650-723-7264 (Voice mail), Statistics 650-723-1185
tibs@stat.stanford.edu
http://www-stat.stanford.edu/~tibs
-----------------------------------------------------------------------
This message was distributed by s-news@wubios.wustl.edu. To unsubscribe
send e-mail to s-news-request@wubios.wustl.edu with the BODY of the
message: unsubscribe s-news