Re: st: parallel processing in Mata

[Jorge Eduardo Pérez Pérez <perez.jorge@ur.edu.co>]

Maybe the solution is to implement a parallelized version of -for-
that only works for loops whose iterations do not depend on results
from previous iterations, an analog to Matlab's -parfor-

This should definitely be in the Stata 12 wishlist.
_______________________
Jorge Eduardo Pérez Pérez



On Thu, Nov 25, 2010 at 9:24 AM, Maarten buis <maartenbuis@yahoo.co.uk> wrote:
>
> --- On Thu, 25/11/10, Mak, Timothy wrote:
> > I've been wanting to ask this question for a while. I don't
> > have many large datasets, but I bought Stata/MP instead of
> > Stata/SE in the hope that it helps me run simulations
> > faster. Does Mata know how to automatically parallelize Mata
> > codes, in particular loops, when it compiles? I can't seem
> > to find information on this anywhere...
>
> I am no expert, but loops seem to me very hard to parallelize.
> The very idea behind a loop is that the second itteration
> happens after the first itteration, the third itteration
> happens after the second itteration, etc. Many applications
> of loops would break if that were not the case. I can see
> why this would not be an issue with a simulation, but how
> is Stata/Mata supposed to know?
>
> Hope this helps,
> Maarten
>
> --------------------------
> Maarten L. Buis
> Institut fuer Soziologie
> Universitaet Tuebingen
> Wilhelmstrasse 36
> 72074 Tuebingen
> Germany
>
> http://www.maartenbuis.nl
> --------------------------
>
>
>
>
> *
> *   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/