st: nlogitrum: flat or discontinuous region

[John Fulton <fulton@ssc.wisc.edu>]

Hi all.  After receiving very generous help from Rich Gates, I still 
can't get
nlogitrum to run properly.  So I'm turning back to statalist.
I have a 3-level model.  One nest is degenerate.  When I run the model, 
I receive the following (new) error:
   could not calculate numerical derivatives
   flat or discontinuous region encountered
   r(430);
What general phenomenon (or phenomena) causes this error?

I can present two facts which may be clues:
1).  When I remove the degenerate nest, nlogitrum runs on the remaining 
data with no errors.
(This entails going to a 2-level model, since the top level has 2 nests, 
one of which is degenerate.)
2).  Nlogit - the original version - runs the 3-level degenerate model 
with no errors.

These clues obviously don't prove anything specific, and I'm not even 
sure if they're indicative of anything at all.  But they would seem to 
narrow the problem down to a difference between nlogitrum and nlogit in 
the way they process my degenerate nest and data.

I'd appreciate any help at all.
Thank you.
John.



*
*   For searches and help try:
*   http://www.stata.com/support/faqs/res/findit.html
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/