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]

st: relax the convergence criteria in ivprobit model


From   xueliansharon <xuelianstata@gmail.com>
To   statalist@hsphsun2.harvard.edu
Subject   st: relax the convergence criteria in ivprobit model
Date   Sat, 21 May 2011 14:28:46 -0700 (PDT)

Hi,

When I run the ivprobit model in Stata, sometimes the ivprobit model doesn't
converge under the default convergence criteria, however, when I relax the
criteria a bit, e.g. relaxing it from 1e-7 to 2e-7, the model converges. And
when I try different sets of IVs, the coefficient estimates are quite
similar. 

Are the coefficients gotten under such relaxed convergence criteria
reliable? 

Another question: in the help file of "maximize", it says that
"nrtolerance(#) specifies the tolerance for the scaled gradient. 
Convergence is declared when g*inv(H)*g' < nrtolerance(). The default is
nrtolerance(1e-5)." However, when I run ivprobit in the reality, I find that
the default value for nrtol is (1e-7), not (1e-5). Any explanation about
this?

Thanks for your response in advance.

Sharon

--
View this message in context: http://statalist.1588530.n2.nabble.com/relax-the-convergence-criteria-in-ivprobit-model-tp6390316p6390316.html
Sent from the Statalist mailing list archive at Nabble.com.
*
*   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