Re: st: collapse if at least X or more obs. per group are non-missing

[Nick Winter <nw53@cornell.edu>]

At 12:19 PM 8/24/2004 -0700, you wrote:
> I'd like to convert a panel data set with annual country obs to one of
> decadal country averages, excluding those decadal averages for which 5 or
> more observations per country are missing within a decade.
>
> My first idea was to use a two step procedure along the lines of:
>
> 1. Run:
> collapse (mean) GDP (count) GDP , by(country decade)
> this should give me a) the decadal averages I want and b) the number of
> non-missing obs used to compute each of these decadal averages.
>
> 2. replace mean_GDP=. if count_GDP<5 (or whatever STATA will call these vars
> saving the means & counting the non-missing obs)
> this should set to missing those decadal averages for which 5 or more
> observations per country were missing within each decade.
It has already been pointed out that you must give different names to your 
new variables with this solution.

Another way to go would be:

        . bysort country decade: generate N = _N
        . collapse GDP if N>5

This would be appropriate, obviously, if you only need the count in order 
to drop the ones you don't want; if you need it for other things then go 
ahead with the original twostep procedure.

Nick WINter


--------------------------------------------------------
Nicholas Winter                           607.255.8819 t
Assistant Professor                       607.255.4530 f
Department of Government              nw53@cornell.edu e
308 White Hall            falcon.arts.cornell.edu/nw53 w
Cornell University
Ithaca, NY 14853-4601 

*
*   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/