RE: st: RE: mat size

["Maarten Buis" <M.Buis@fsw.vu.nl>]

Ines:
With xtreg you are already taking care of the 
"codempr-specific" intercepts. As a result you don't 
need to make them by adding all those dummies. The 
following will suffice -xtreg perform , fe i(codempr)-.
In essence -xtreg , fe- uses a more numerically stable 
way of doing what you wanted to do with adding all those 
dummies. If you add all the dummies you are 
estimating a regression with 1231 variables, which 
isn't pretty. 

HTH,
Maarten

-----------------------------------------
Maarten L. Buis
Department of Social Research Methodology 
Vrije Universiteit Amsterdam 
Boelelaan 1081 
1081 HV Amsterdam 
The Netherlands

visiting adress:
Buitenveldertselaan 3 (Metropolitan), room Z434 

+31 20 5986715

http://home.fsw.vu.nl/m.buis/
-----------------------------------------

-----Original Message-----
From: owner-statalist@hsphsun2.harvard.edu [mailto:owner-statalist@hsphsun2.harvard.edu]On Behalf Of irodriguez
Sent: woensdag 1 november 2006 14:03
To: statalist@hsphsun2.harvard.edu
Subject: Re: st: RE: mat size

I�ve tried this:
    xi: xtreg perform i.codempr, fe i(codempr)

But the outcome is similar:
 i.codempr         _Icodempr_1-1203    (naturally coded; _Icodempr_1
omitted)
 matsize too small
 You have attempted to create a matrix with more than 800 rows or columns
 or to fit a model with more than 800
 variables plus ancillary parameters.  You need to increase matsize using
the
 set matsize command; see help matsize.
 r(908);

What I�m doing wrong?
thanks a lot

-----Original Message-----
From: "Maarten Buis" <M.Buis@fsw.vu.nl>
To: <statalist@hsphsun2.harvard.edu>
Date: Wed, 1 Nov 2006 13:45:39 +0100
Subject: st: RE: mat size

> Ines:
> Apparently you are trying to estimate a fixed effects regression by
> adding
> dummies. A better way of doing the fixed effects regression is -xtreg-
> with
> the fe option. see: -help xtreg- and the entry for -xtreg- in the xt
> manual.
>
> HTH,
> Maarten
>
> -----------------------------------------
> Maarten L. Buis
> Department of Social Research Methodology
> Vrije Universiteit Amsterdam
> Boelelaan 1081
> 1081 HV Amsterdam
> The Netherlands
>
> visiting adress:
> Buitenveldertselaan 3 (Metropolitan), room Z434
>
> +31 20 5986715
>
> http://home.fsw.vu.nl/m.buis/
> -----------------------------------------
>
> -----Original Message-----
> From: owner-statalist@hsphsun2.harvard.edu
> [mailto:owner-statalist@hsphsun2.harvard.edu]On Behalf Of irodriguez
> Sent: woensdag 1 november 2006 13:33
> To: statalist@hsphsun2.harvard.edu
> Subject: st: mat size
>
> Dear list-members,
>
> I am trying to run a OLS regression with panel data (8240 observations)
> The explaining variable have 1203 categories
>
> I estimate the following model:
>     set mem 50m
>     set matsize 800
>     xi: reg perform i.codempr
> but the outcome is:
>
> i.codempr         _Icodempr_1-1203    (naturally coded; _Icodempr_1
> omitted)
> matsize too small
>     You have attempted to create a matrix with more than 800 rows or
> columns
> or to fit a model with more than 800
>     variables plus ancillary parameters.  You need to increase matsize
> using
> the set matsize command; see help matsize.
> r(908);
>
>
> *
> *   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/


*
*   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/

*
*   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/