st: Break if no convergence after certain time

[<S.Jenkins@lse.ac.uk>]

----------------------------------------------------------------------

Date: Sat, 9 Jul 2011 11:49:17 +0200
From: Stefan Pichler <stefan__pichler@hotmail.com>
Subject: st: Break if no convergence after certain time

Dear statalisters,

I am running several large maximization problems in Stata. Is there a
way to interrupt Stata and proceed with the next maximization if no
convergence has been reached after a certain time?
Thanks for your help.
----------------------------------------------------------------------

I don't know how to do interruptions after a specific amount of elapsed
time; I doubt it's possible.

However, a feasible alternative is to use the -iterate(#)- option in
conjunction with the value of e(converged). You might then also use
-estimates save-, and later restart the problem using -estimates use-
and -from(init_specs)- option.  See -help maximize- and -help
estimates-.

That is, suppose well-behaved problems take about 10 iterations (say).
Set -iterate(100)- (say), so that well-behaved problems converge, and
poorly-behaved ones won't have after 100 iterations -- but the e()
values will be set whatever the case. You can check post-hoc whether
there is convergence from e(converged) which equals 1 if converged, 0
otherwise.

Stephen
-------------------------------------
Professor Stephen P. Jenkins  <s.jenkins@lse.ac.uk>
Department of Social Policy and STICERD
London School of Economics and Political Science
Houghton Street, London WC2A 2AE, U.K.
Tel. +44 (0)20 7955 6527
Survival Analysis using Stata:
http://www.iser.essex.ac.uk/survival-analysis
Downloadable papers and software: http://ideas.repec.org/e/pje7.html


Please access the attached hyperlink for an important electronic communications disclaimer: http://lse.ac.uk/emailDisclaimer

*
*   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/