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

st: RE: postfile problem


From   "Nick Cox" <n.j.cox@durham.ac.uk>
To   <statalist@hsphsun2.harvard.edu>
Subject   st: RE: postfile problem
Date   Thu, 6 Feb 2003 11:35:45 -0000

Paul Silcocks BM MSc FRCPath FFPHM CStat
> 
> I'm trying to estimate power of a non-linear regression by 
> simulation.
> There are three parameters of interest and I can generate 
> samples under H0
> (no trend) and H1 (particular model of interest).
> 
> The problem is how to write the parameter estimates under 
> H0 and H1 to a
> file.  In the code below, the estimation is performed twice 
> - under H0 and
> H1.  What I would like to do is post the 6 estimates in one 
> go to the output
> file, rather than output to two separate files which would 
> be merged.
> 
> forvalues i=1/5000 {
>           /* evaluate when H0 true */
>               replace err = 0 + 3.5*invnorm(uniform())
>               replace tempy = 0 +err
>                     nl hyp tempy dose, iterate(30) eps(0.0001)
> 
>  /* How do I save parameter estimates from H0 ? */
> 
>                  /* evaluate when H1 true */
>                 replace tempy = yh +err
>                     nl hyp tempy dose, iterate(30) eps(0.0001)
> 
>    /* here I save the results from estimate given H1 */
> 
>              post `powerdat'  (_b[B0])  (_b[B1])  (e(r2_a))
> 
> 
> Any thoughts anybody?  Ideally the last statement would be like:
>          
>            post `powerdat' (_b[B0])  (_b[B1])  (e(r2_a))  
> (_b[B0])  (_b[B1])
> (e(r2_a))
> 
>  Where the first 3 are from the estimate under H0 and the 
> second 3 are from
> H1
> 

Put those first estimates somewhere where they won't 
be overwritten before being posted, e.g. 

scalar B0 = _b[B0] 
scalar B1 = _b[B1] 
scalar r2_a = e(r2_a) 

and then put those names in a single call to -postfile-. 

Nick 
n.j.cox@durham.ac.uk 

*
*   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