Many thanks to Charla Lambert from Statsci for confirming that there is a
bug in SPlus 2000's predict.discrim() function. Using the period '.'
syntax to define a discrim() object causes the predict.discrim() function
to fail whenever the 'newdata=' argument is used. The only known
workaround at this time is to explicitly list all the dependent variables
in the original call to discrim(). It isn't a perfect solution (especially
with a large number of dependent variables), but at least it works.
Thanks Charla.
=====================
Dr. Marc R. Feldesman
Professor and Chairman
Anthropology Department
Portland State University
1721 SW Broadway
Portland, Oregon 97201
email: feldesmanm@pdx.edu
phone: 503-725-3081
fax: 503-725-3905
http://odin.cc.pdx.edu/~h1mf
======================
-----------------------------------------------------------------------
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
|