Stata The Stata listserver
[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

Re: st: xtrchh2/xtrchh


From   "Brian P. Poi (StataCorp LP)" <bpoi@stata.com>
To   statalist@hsphsun2.harvard.edu
Subject   Re: st: xtrchh2/xtrchh
Date   Wed, 9 Jun 2004 10:29:19 -0500 (CDT)

On Wed, 9 Jun 2004, Diego Rei wrote:

> Hello to all the Statalisters, I am trying to run the commands
> xtrchh/xtrchh2 on my models and I always get the following message:  
> "602 obs. dropped (panels too small)" Does anybody know why this is
> happening? Suspecting that it might depend on the rather small number of
> cross sections relative to parameters to be estimated, I eliminated some
> of the covariates and obtained : "matrix has missing values"
> which is indeed true. I just would ask for a confirmation, is the xtrchh 
> command running with balanced panels only?Is the problem with panels too 
> small related to the number of panels with respect to the parameters or 
> to the time series dimension of the cross sections, or 
> to something else?
> Thanks for any help.
> Have a good day
> 


Dear Diego,

In a random-coefficients model, the number of observations in each 
panel must be greater than the number of regressors (including the 
constant).  The first step in fitting Swamy's RC model is to perform 
OLS regression on each panel separately, and if the condition I just 
mentioned is not met, OLS will not work.  Panels that do not meet that
requirement will get dropped.

If you still have problems, feel free to email me your dataset 
privately and the exact syntax you are using.

Hope this helps.

   -- Brian Poi
   -- bpoi@stata.com

*
*   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/



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