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 on April 23, and its replacement, statalist.org is already up and running.


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

Re: st: RE: r(1000) error when running xi


From   A Loumiotis <antonis.loumiotis@gmail.com>
To   statalist@hsphsun2.harvard.edu
Subject   Re: st: RE: r(1000) error when running xi
Date   Tue, 1 Nov 2011 09:08:27 +0200

The problem is not limited to -xi- of -ice-.  For example the
following code stops with a r(920) error when the length of local test
reaches 165198.

local test
forvalues i=1/166000 {
	local test "`test' `i'"
	local ctest : length local test
	di "`ctest'"
}

...
165180
165186
165192
165198
too many macros
r(920);



On Mon, Oct 31, 2011 at 2:55 PM, Nick Cox <n.j.cox@durham.ac.uk> wrote:
> I do have a wild guess, which is that the limit on macro length may be affected by a -version- declaration somewhere along the computational path.
>
> Nick
> n.j.cox@durham.ac.uk
>
>
> -----Original Message-----
> From: owner-statalist@hsphsun2.harvard.edu [mailto:owner-statalist@hsphsun2.harvard.edu] On Behalf Of Nick Cox
> Sent: 31 October 2011 12:40
> To: 'statalist@hsphsun2.harvard.edu'
> Subject: RE: st: RE: r(1000) error when running xi
>
> I see. There is a division of labour here: if the problem appears to be in -ice- then tech-support is likely to refer you to its author.
>
> I have no guesses what is going on. I try never to estimate more about seven parameters at a time.
>
> Nick
> n.j.cox@durham.ac.uk
>
> A Loumiotis
>
> Thanks Nick.
> I will contact tech support and I will take your advise from now on
> and clone official stata commands before I modify them.
>
> The problem occurs when I use -ice- which does not accept factor
> variables (and the -ice- code uses xi).
>
> SJ-9-3  st0067_4  . Mult. imp.: update of ice, with emphasis on cat. variables
>        (help ice, uvis if installed) . . . . . . . . . . . . . . . P. Royston
>        Q3/09   SJ 9(3):466--477
>        update of ice package with emphasis on categorical
>        variables; clarifies relationship between ice and mi
>
>
> On Mon, Oct 31, 2011 at 2:09 PM, Nick Cox <n.j.cox@durham.ac.uk> wrote:
>> Also, -xi:- is not needed in Stata 11, given the introduction of factor variables.
>>
>> Nick
>> n.j.cox@durham.ac.uk
>>
>> A Loumiotis
>>
>>
>> I get the following error when I run xi:
>>
>> system limit exceeded - see manual
>> r(1000);
>>
>> I modified a bit xi and measured the length of the local orig and
>> found out that the error occurs when the macro length of orig exceeds
>> 165198.  I'm surprised because I'm not using Stata/IC where the macro
>> limit is 165,200 but I'm using Stata/MP where the macro limit is
>> 1,081,511.  Do you know why this is happening?
>>
>> My version of Stata is the following:
>>
>> .about
>> Stata/MP 11.2 for Windows (32-bit)
>> Born 01 Sep 2011
>> Copyright 1985-2009 StataCorp LP
>>
>> Total physical memory:     2097151 KB
>> Available physical memory: 1726136 KB
>>
>
> *
> *   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