In response to my recommendation to use -mlogpred-, Matthew R. Polisson
<[email protected]> writes:
> Thanks very much for your help. The mlogpred command works quite nicely. Is
> there a similar command which follows logit? I am assuming that predictnl
> works after logit, although I can imagine running into similar iteration
> problems.
I don't know of an equivalent program for -logit-, however, you are unlikely
to get convergence problems using -predictnl- after -logit-, since the
calculation for predicted probabilities after -logit- is fairly sturdy. If
you do run into problems, I'd be most interested to have a look at your data
and model.
On another note, you can use -mlogit- to fit a -logit- model, e.g. with the
auto dataset
. mlogit foreign turn displacement, base(0)
vs.
. logit foreign turn displacement
and thus still use -mlogpred-. Yeah, using -mlogit- for -logit- is overdoing
it a bit, but it works.
--Bobby
[email protected]
*
* 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/
