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: Bootstrap got stuck.


From   Stas Kolenikov <skolenik@gmail.com>
To   statalist@hsphsun2.harvard.edu
Subject   Re: st: Bootstrap got stuck.
Date   Fri, 12 Aug 2011 15:57:01 -0400

On Fri, Aug 12, 2011 at 3:13 PM, Xu, Haiyong <HXu@mednet.ucla.edu> wrote:
> When I bootstrap a model, e.g zinb, in Stata, it often gets stuck. I think it is because in some iteration the model can't get fitted because when I let it run noisily it will show me "not concave" in the model fitting part. I try to set different seeds but it doesn't help. Is it because my model is not appropriate? Or is there anything else I can try to avoid this, such as setting initial values or saving some successful iterations and calculate the confidence interval manually? Any advice is appreciated. Thank you very much!

This is a nature of bootstrap -- conceptually, it is possible to
obtain a data set with a single unique observation propagated _N
times, in which no model would make any sense. To trim these awkward
samples, you might want to explicitly specify the number of ML
iterations in the program that is being called with -iter(50)-, say.
Then upon reaching this number, -zinb- or whatever you work with will
declare that convergence was not achieved ( e(converged)==0 ), and
-bootstrap- will omit this repetition marking it with an "x" in the
dotted output.

-- 
Stas Kolenikov, also found at http://stas.kolenikov.name
Small print: I use this email account for mailing lists only.

*
*   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