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: SAS vs STATA : why is xtlogit SO slow ?


From   Amir Sariaslan <amir.sariaslan@gmail.com>
To   statalist@hsphsun2.harvard.edu
Subject   Re: st: SAS vs STATA : why is xtlogit SO slow ?
Date   Thu, 2 Feb 2012 16:35:11 +0100

Francesco,

Try -clogit- and see if it works. Otherwise, try running your model
with a subset of your data where you select about 5-10 measurement
points per individual to see if it works. The additional data would
probably only change the estimates marginally anyway.

Amir

On Thu, Feb 2, 2012 at 10:52 AM,  <k7br@gmx.fr> wrote:
> Dear Amir,
>
> The STRATA option in proc logistic should indeed account for
> individual fixed effects.. (see here
> http://www.ats.ucla.edu/stat/sas/faq/conditional_logit_sas.htm)
> I am very sad not being able to use XTLOGIT with my database... (my
> computer has 12go RAM, proc i7)
>
> Any more ideas ?
> Thx
>
> On 2 February 2012 01:25, Amir Sariaslan <amir.sariaslan@gmail.com> wrote:
>> Francesco,
>>
>> If I'm not mistaken, the SAS syntax generates cluster-robust standard
>> errors - not fixed effects logistic regression. The equivalent syntax
>> in Stata would thus be: -logit depvar indepvar, cluster(clustervar)
>> or-. For further information as to the differences between the models,
>> see: http://repec.org/usug2007/crse.pdf
>>
>> Amir
>>
>> On Wed, Feb 1, 2012 at 6:58 PM,  <k7br@gmx.fr> wrote:
>>> Dear Statalist,
>>>
>>> I have a very large panel dataset (about 7mo observations, 70 000
>>> individuals, 50 points on average per individual) and I tried
>>> desperately to estimate a fixed effect logit using : xtlogit, fe with
>>> Stata...
>>> Unfortunately the log likelihood never converge and stops at iteration
>>> 1 or 2 saying that the matrix is not semi definite positive...
>>> (Iteration 0:   log likelihood =    -1.#INF , etc)
>>>
>>> However when I try with SAS (using proc Logistic with the strata
>>> option) the computation is quite long (more than one hour) but at
>>> least I can see the results...
>>>
>>> My question is : how is that possible ? I believed that storing the
>>> database into the RAM (as STATA do) would speed computations... and I
>>> would love to use xtlogit (or clogit) instead of using SAS...
>>>
>>> any idea ?
>>>
>>> Many thanks,
>>>
>>> Francesco
>>> *
>>> *   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/
>
> *
> *   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