Bookmark and Share

Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.


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

Re: st: Benchmarks for Stata


From   Stas Kolenikov <[email protected]>
To   [email protected]
Subject   Re: st: Benchmarks for Stata
Date   Thu, 9 Dec 2010 17:40:01 -0600

I have some estimates of computational complexity in my Stata Journal
paper on -confa-, if that helps. These are comparative figures: if you
increase the sample size by a factor of two, or add another indicator
to each factor, how much longer would it take to estimate the model?
But you are probably after an absolute index of performance.

I wouldn't be at all surprised that a modern laptop can ran as fast as
a slightly dated server. Especially if that server is already loaded
with some computing, and/or you don't have sufficient priority in the
computing queues.

On Thu, Dec 9, 2010 at 5:32 PM, Dimitriy V. Masterov <[email protected]> wrote:
> Are there any standardized tests that one can run to measure Stata's
> performance? I am becoming convinced that my current setup is slower
> than it could be, but I don't know any way of confirming that this is
> not a case of unrealistic expectations. I've seen some suggestions for
> benchmarking using gllamm, but I don't know enough about gllamm models
> to know which are parallelized and to what degree. I guess I don't
> even know what to expect.
>
> I am running Stata MP on a 64-bit Windows server through remote
> desktop, and it seems to be only somewhat faster than MP on my laptop.
> Any advice about how to establish expectations or how to diagnose what
> is wrong?
>



-- 
Stas Kolenikov, also found at http://stas.kolenikov.name
Small print: I use this email account for mailing lists only.
*
*   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 1996–2018 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   Site index