Re: st: geting e-class returns in outreg2 for multiple models

["Miller, Daniel P" <dpmiller@bu.edu>]

Hi Doug - glad it worked partially. Perhaps the solution is to loop the regression command, store the proper estimates in the loop and then use a single outreg2 command? 

Sent from my iPhone

On Jul 8, 2011, at 10:52 AM, "Doug Hess" <douglasrhess@gmail.com> wrote:

> Thanks, Dan. The append did work great for putting multiple models
> side-by-side, which is very cool. But it still gave each model the
> same LL.
> 
> Maybe I created the outreg2 loop incorrectly? I'm just figuring
> looping out. Below is what I wrote; it comes after running and storing
> by "est store" a dozen models. Is the issue that I need the LL to be
> stored perhaps as a local macro call "LL1" , "LL2" etc. after each
> model and then called up in the "addstat()" option with LL`i' ? Or
> should I do an "outreg, append" after each model is run and forget the
> loop? I was hoping for shorter code, but maybe pithiness isn't the
> right virture here.
> 
> forvalues i=1/12 {
> outreg2 [fullmodellogit`i'] using July8fullmodel, sideway  ///
>        stats(coef se pval aster) addstat(LL, `e(ll)') noparen dec(3) excel append
> }
>    
> Thank you!
> 
> -Doug
> 
> Date: Thu, 7 Jul 2011 19:56:09 -0400
> From: "Miller, Daniel P" <dpmiller@bu.edu>
> Subject: Re: st: geting e-class returns in outreg2 for multiple models
> 
> Hi Doug, I can't speak to the repeat of the LL, but my experience has
> been that append stores the results side by side when you use the
> excel option.
> 
> Dan
> 
> On Jul 7, 2011, at 6:10 PM, Doug Hess wrote:
> 
>> Hi, all.
>> 
>> I am trying to produce a single file with multiple regressions
>> side-by-side, which outreg2 does wonderfully. However, when I run the
>> outreg2 code below after running the regressions and storing after
>> each with -est store- using sequential names, the LL that is presented
>> underneath each model is the same one (the one from the last
>> regression run).  Should I also be storing the e-class returns somehow
>> and then calling them up? Or will this require a -foreach- loop, which
>> loses the side by side model output (I assume -append- puts things at
>> the bottom?).
>> 
>> outreg2 [model*] using July7models, sideway stats(coef se pval aster)
>> addstat(LL, `e(ll)') noparen dec(3) excel word replace
>> 
>> Thanks for any ideas.
>> 
>> Doug
> *
> *   For searches and help try:
> *   http://www.stata.com/help.cgi?search
> *   http://www.stata.com/support/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/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/