st: Re: Question about programming using bootstrap

["R.E. De Hoyos" <redeho@hotmail.com>]

Alejandro,

You can estimate the empirical standard errors the scalars your self using 
the bsample command.

I haven't try the following code but you can work along this lines:

local K=100 //Number of replications
local k=1
preserve
while `k'<=`K' {
   bsample
   foreach var of local varlist {
       tempname mg`var'
       scalar mg`var' = sg`var' - s`var'
       tempname mg`var'`k'
       matrix `mg`var'`k''=mg`var'
       if `k'==1 {
           matrix `mg`var'=`mg`var'`k''
       }
       if `k'>1 {
           matrix `mg`var''=`mg`var''+`mg`var'`k''
       }
       restore, preserve
       local k=`k'+1
   }
   local k=1
   while `k'<=`K' {
       if `k'==1 {
           matrix se_mg`var'=(`mg`var'`k'' - 
(`mg`var''/`K'))'*(`mg`var'`k'' - (`mg`var''/`K'))
       }
       if `k'>1 {
           matrix se_mg`var'=se_mg`var'+(`mg`var'`k'' - (`mg`var''/`K'))'* 
///
           (`mg`var'`k'' - (`mg`var''/`K'))
       }
       local k=`k'+1
   }
   matrix se_mg`var'=se_mg`var'/(`K'-1)
   restore
}

I hope this helps,

Rafa
________________________
R.E. De Hoyos
Faculty of Economics
University of Cambridge
CB3 9DE, UK
www.econ.cam.ac.uk/phd/red29/

----- Original Message ----- 
From: "feldman" <feldman@primal.ucdavis.edu>
To: <statalist@hsphsun2.harvard.edu>
Sent: Thursday, September 15, 2005 3:04 PM
Subject: st: Question about programming using bootstrap



> Statalisters,
> I have a question about how to bootstrap a set of variables
> using a single
> command. More specifically what I am interested in doing is
> the following.
> I create a program that estimates a scalar for a set of
> variables defined
> in varlist. I then want to bootstrap that program to obtain
> the standard
> errors of each one of the scalars. Nevertheless, I can't
> find a way to tell
> Stata to do this using a single command for bootstrap. Below
> is the
> relevant part of the code that I am using. I get the
> bootstrapped standard
> errors but in a very inefficient way because Stata repeats
> the bs command
> for each one of the variables in varlist and that can be
> extremely time
> consuming. I want to do this with a single bs command.
>
> Thanks,
> Alejandro
>
> P.S. My apologies if this message is repeated. I sent it
> before from another email account (twice) but it didn't
> went through after more than 12 hours.
> _______
>
> program define desc
> .
> .
> .
> foreach var of local varlist {
> scalar mg`var' = sg`var' - s`var'
> }
> .
> .
> .
>
> foreach var of local varlist {
> set seed  5432198
> bs "desc `varlist' " mg`var'=mg`var' , rep(`rep')
> level(`level')
> }
> _______________
>
>
> Alejandro Lopez-Feldman
> Ph.D. Candidate
> UC DAVIS
> Agricultural and Resource Economics
>
> *
> *   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/
*
*   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/