--- "Kelchtermans, Stijn" <Stijn.Kelchtermans@e...> wrote:
> based on the parameter estimates, get 1000 draws from the
> multivariate normal distribution to generate 1000 alternative
> parameter values. Using these values, I calculate 1000 choice
> probabilities. I then drop 2,5% of the values in each tail to get
> the 95% CI for the choice probabilities.
>
> I can't find Stata commands that allow to automate this procedure.
Have a look at -simulate-, something like this should do what you
want:
program define sim, rclass
drop _all
whatever commands needed to generate one "new" dataset
whatever commands needed to estimate parameters
whatever commands needed to compute probability
return scalar prob = probability
end
simulate sim prob = r(prob), dots reps(1000)
sum prob, detail
Since you are repeating these commands a thousand times it is well
worth to use build in commands instead of ado commands wherever
reasonably possible, since the build in commands are a lot faster.
However, it will always remain a tradeoff between the time it takes
to code your do-files (and later understand and revise your do-file)
and the time it takes to run your do-file.
Maarten
*
* 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/