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: Time complexity of -scobit-
From
Stas Kolenikov <[email protected]>
To
"[email protected]" <[email protected]>
Subject
Re: st: Time complexity of -scobit-
Date
Tue, 22 Oct 2013 16:46:29 -0500
I estimated the computational complexity for my confirmatory factor
analysis module, -confa-, see
http://www.stata-journal.com/article.html?article=st0169. I also found
computing time to be roughly linear in N, and of power 2.4 in model
complexity (as expected -- this is the exponent on the fastest
available matrix inversion algorithm). For complicated likelihoods,
convergence is slower in smaller samples as the likelihood surfaces
are just more weird, with non-concave regions and such.
Hint: -help timer-.
-- Stas Kolenikov, PhD, PStat (ASA, SSC)
-- Senior Survey Statistician, Abt SRBI
-- Opinions stated in this email are mine only, and do not reflect the
position of my employer
-- http://stas.kolenikov.name
On Tue, Oct 22, 2013 at 3:33 PM, Sergiy Radyakin <[email protected]> wrote:
> Zachary, I don't know about -scobit-, but ordinary logit (not ordered
> logit) is roughly linear in N:
>
> 1000-->.491
> 10000-->4.861
> 20000-->9.497
> 30000-->14.724
> 40000-->18.99
> 50000-->24.267
>
> (Stata MP 12 restricted to use 1 CPU only).
>
> Hope this helps, Sergiy Radyakin
>
>
>
>
>
> On Tue, Oct 22, 2013 at 3:11 PM, Zachary Neal <[email protected]> wrote:
>> I am working on an algorithm that involves estimation using -scobit-
>> and am trying to determine the time complexity of the algorithm. Are
>> there any estimates of the time complexity of -scobit-, or even of
>> ordinary -logit-, as a function of N? Given the ML estimation, is
>> finding the time complexity as a function of N even possible?
>>
>> Thanks,
>> Zachary
>>
>> --
>> Zachary Neal, PhD
>> Department of Sociology
>> Michigan State University
>> [email protected]
>> http://www.msu.edu/~zpneal
>> *
>> * For searches and help try:
>> * http://www.stata.com/help.cgi?search
>> * http://www.stata.com/support/faqs/resources/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/faqs/resources/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/faqs/resources/statalist-faq/
* http://www.ats.ucla.edu/stat/stata/