Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
From | Ian Breunig <ianbreunig@gmail.com> |
To | statalist@hsphsun2.harvard.edu |
Subject | st: missing output in ml |
Date | Wed, 7 Apr 2010 12:09:55 -0600 |
Statlist, After running ml on a system of equations with a user written likelihood function, I find that sometimes the results show missing values ('.') for the standard errors and related results on some of the estimated '_cons', and sometimes for other constants being estimated in the model. The e(v) matrix shows that the variance matrix has a zero across the row for this parameter. The coefficient itself is estimated. Can anyone tell me what this means? What is going wrong or what does this imply? (The same results occur when I use pweights and svyset with pweights specified). I've searched through the Gould et al. book, and searched the web already but couldn't find an answer. Thanks. Ian Breunig * * For searches and help try: * http://www.stata.com/help.cgi?search * http://www.stata.com/support/statalist/faq * http://www.ats.ucla.edu/stat/stata/