Re: st: Keep/Drop Observations for Top/Bottom X%

["Justina Fischer" <JAVFischer@gmx.de>]

Hi Maarten,

exactly, when working with really big datafiles (e.g. > 100'000 obs.) and estimating a non-linear model dropping the obs not in use saves heavily on computation time, or might even estimations become manageable in the first place (e.g. when estimating interaction effects). 

Justina


-------- Original-Nachricht --------
> Datum: Thu, 11 Oct 2012 12:07:43 +0200
> Von: Maarten Buis <maartenlbuis@gmail.com>
> An: statalist@hsphsun2.harvard.edu
> Betreff: Re: st: Keep/Drop Observations for Top/Bottom X%

> On Thu, Oct 11, 2012 at 11:54 AM, Justina Fischer  wrote:
> > in principle you might be right.
> >
> > However, for reasons of practicability it is sometimes recommendable for
> subset analysis to simply upload the full data and drop a part rather than
> working with an 'if' restriction throughout all regressions.
> 
> It is largely a matter of style. I like the principle of keeping your
> data as much as possible intact, and I thus prefer the -if- route over
> the -keep- route. Using if selections throughout my analysis has
> become natural for me, and even desirable as a constant reminder of
> which sub-sample I am working on. The main reason why I sometimes
> deviate from that default is when the data is so large (e.g. Census
> data) that it becomes unmanageable.
> 
> -- Maarten
> 
> ---------------------------------
> Maarten L. Buis
> WZB
> Reichpietschufer 50
> 10785 Berlin
> Germany
> 
> http://www.maartenbuis.nl
> ---------------------------------
> *
> *   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/