Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: st: initial values not feasible
From
Nick Cox <[email protected]>
To
[email protected]
Subject
Re: st: initial values not feasible
Date
Sun, 2 Dec 2012 18:11:45 +0000
-xtmpg- is from SJ 7-2. Please remember...
I am not clear what kind of help you expect us to be able to give.
Perhaps the parameter estimates for the model that converged would be
reasonable starting points for trying again with the model that
didn't. Either way, your possibilities clearly range from trying
non-default starting values to concluding that the first model is
inappropriate for your data.
There is little white magic with such models. They're always written
up with examples that work well -- I've done precisely the same in
other territories -- but they won't work if the generating process is
something different. Often one needs to take a hard look at how many
parameters are being estimated with how much data (how many panels,
how long).
Nick
On Sun, Dec 2, 2012 at 6:01 PM, urbain thierry YOGO <[email protected]> wrote
> I am trying to run a pooling mean group estimation using xtpmg. My
> equation looks like this:
> xtpmg d.ltgdp d.(ldep lrev dte), ec(ec) lr( ldep lrev dte), pmg replace
> however stata gives me back the error 1400 "initial values not feasible).
> But whent i run the model only with the first two variables (ltgdp
> ldep) stata gives me the right output.
*
* 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/