Bookmark and Share

Notice: On March 31, it was announced that Statalist is moving from an email list to a forum. The old list will shut down on April 23, and its replacement, statalist.org is already up and running.


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: st: Probit regression does not work, convergence not achieved


From   Nick Cox <njcoxstata@gmail.com>
To   "statalist@hsphsun2.harvard.edu" <statalist@hsphsun2.harvard.edu>
Subject   Re: st: Probit regression does not work, convergence not achieved
Date   Mon, 3 Jun 2013 22:25:53 +0100

The best guess is that you tried to fit too complicated a model, but
there are other possibilities.

There is no information here on

1. Sample size.

2. Numbers of 0s and 1s.

3. Structure of predictors, including interaction terms. (Often people
try to fit a model with many predictors to a rather small dataset.)

4. Distribution of predictors, including any highly skewed variables
or variables with outliers.

Nick
njcoxstata@gmail.com

Teri ( Xilin  Zhou)  [edited]

I'm trying to run probit regressions for a binary outcome model. I
have two binary outcomes, A and B. B is a sub-population of A.

When I run -probit- for outcome A, it went well, but when I try
outcome B, the estimation went on over-night, and after thousands of
iterations of the ML, Stata reported "convergence not achieved,
r(430)".

I tried to use alternatives, such as -logit-, or -glm,  f(b) l(p)-,
but none of them gave me results.

I'm also sure that I don't have a separation problem of my data. Can
somebody help me out?
*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/faqs/resources/statalist-faq/
*   http://www.ats.ucla.edu/stat/stata/


© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   Site index