Bookmark and Share

Notice: On March 31, it was announced that Statalist is moving from an email list to a forum. The old list will shut down at the end of May, and its replacement, statalist.org is already up and running.


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: st: how to reduce running time and data size


From   Randall Lewis <randallalewis@gmail.com>
To   statalist@hsphsun2.harvard.edu
Subject   Re: st: how to reduce running time and data size
Date   Sat, 24 Apr 2010 04:07:56 -0700 (PDT)

Discretizing the support and using fweights might speed up the process. (Obviously, you'll just have to create a new weighting vector--if you can reduce the 10,000 obs down to a few hundred, you'll speed it up by quite a bit.) But not knowing exactly what your simulation is for, this might be impossible or nonsensical.

--Randall


----- Original Message -----
From: "Abdel Rahmen El Lahga" <rahmen.lahga@gmail.com>
To: statalist@hsphsun2.harvard.edu
Sent: Friday, April 23, 2010 1:34:17 PM GMT -08:00 US/Canada Pacific
Subject: Re: st: how to reduce running time and data size

Also it seems that your sample size is excessivly large (10000). For a
monte carlo study I think that you will be ok with 3000 or even 2000
observation
HTH
AbdelRahmen

2010/4/23 Maarten buis <maartenbuis@yahoo.co.uk>:
> --- On Fri, 23/4/10, sun samn wrote:
>>    I am running simulation about a GMM in MATA; the
>> matrices are only involved Within MATA. My sample size is
>> 10000 and simulate for 1000 times, which makes the running
>> like forever. Plus, it seems like the memory is not big
>> enough, Is there any way I can do to improve this situation?
>
> The simple answer is use fewer observations or fewer iterations.
> More specific advise is hard to give, as we haven't seen your
> code. The are some general comments possible. You could try
> adding starting values if you have good ones, as is often the
> case in a simulation study. You could specify the
> -derivative()- option, but that can mean a lot time spent on
> programming and debugging, so if the time gain you receive from
> the faster program will outweight the time lost programming is
> an open quesiton.
>
> -- 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/
>



-- 
AbdelRahmen El Lahga

*
*   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/


© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   Site index