When Stas says likelihood, I think
he means prior probability!
Nick
[email protected]
Stas Kolenikov
> Stata Corp folks should be able to answer that question best, but my
> impression is that your best option with C is to translate your stuff
> into Stata plugins, see http://www.stata.com/plugins/. The best Stata
> optimizer is by far -ml-, but it only has that specific idea in mind:
> you have a likelihood, and you want to find your MLEs. It is very well
> tuned to that, but because of its universal character (in terms of
> being able to maximize any weird likelihood), it is relatively slow.
> Other optimizers (at least those I've seen) seemed to have a rather
> amateur flavor to me. The stuff that's not in C, and that you cannot
> put into plugins, you would have to rewrite in Mata; as long as the
> latter is a matrix language, it will be way easier to transfer your R
> programs into than original Stata. You may have to write your own code
> for all three pieces. If the FAQs and statalist archives don't give
> you anything serious, then with 98% likelihood it is not there.
>
> On 12/3/05, Jens Hainmueller <[email protected]> wrote:
> > Hi Statalisters,
> >
> > we are trying to translate a package that we have written
> in R and C into
> > Stata. Our method requires several steps of nested optimization.
> > Unfortunatly, I am not very familiar with the several
> optimizers provided in
> > Stata, so I would be grateful if somebody could point me to
> the Stata
> > functions for the following problems (I could not find this
> info in the
> > archives nor the FAQs, sorry if I missed it there):
> >
> > 1. bound constrained optimization (for arbitrary loss functions)
> >
> > 2. quadratic programming
> >
> > 3. genetic optimization
> >
> > Also, is it possible to combine Stata with other engines such C?
> >
> > Thank you very much for your help! I am using Stata 9.1.
*
* 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/
