Re: st: Problem with xtregfem in panel data

[Nick Cox <njcoxstata@gmail.com>]

You are asked to explain _where_ user-written programs come from.

-xtregfem- is a program from SSC by Emad Abd Elmessih Shehata, who is
not I think active on Statalist.

On the face of it, you have the scope to create a matrix of that size.
-xtregfem-, although for some reason stated to require Stata 11, makes
no use of Mata, but heavy use of Stata's matrix language.

For that reason, you should probably

1. get a trace of the program and find out exactly where the program
fails using -set trace on-

2. contact the author directly

You might need to send him a copy of your dataset for his testing.


Nick
njcoxstata@gmail.com


On 10 February 2014 18:12, Rolf Lyneborg Lund <starpen@mac.com> wrote:
> Hello statalisters.
>
> I've been trying to run a userwritten program to do a ridge and weighted fixed effects panel data regression called xtregfem but I can't get it to work. What I do is:
>
> *******example*******
>
> xtregfem hea gdpln postin labforfem, id(country) it(year)
>
>
> ******example end*******
>
> But I get error:
>
> matsize too small to create a [480,1] matrix
>
>
> I've tried to set the matsize higher (since I run STATA I/C, the limit is 800), but I get the same error. Is there any way to do ridged panel data regression without this error popping up? And do I really need to upgrade to run this regression?
>
> Best
> -Rolf Lyneborg Lund
> *
> *   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/
*
*   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/