--- Altay Turk wrote:
> I have been doing frontier regression analysis on individual oil
> reservoirs.
>
> After tsset the data i run (all in log form)
>
> frontier y x z
>
> which runs the technical effiiciency term as half-normal
> distribution.
>
> I get the results after a few iterations.
>
> However, whenever i indicate exponential distribution for the
> technical efficiency term as:
>
> frontier x y z, distribution(exponential)
>
> The log-likelihood stabilizes to a value and keeps on iterating with
> "backed up" message.
It is better to reply to the list instead of to individuals as is
discussed in the statalist FAQ (point 5). This is a good example, as I
know nothing about the specifics of -frontier-, so someone else on the
list might give much better advise. Generaly, this behavior means that
Stata is having trouble finding the maximum of the likelihood function.
One thing you can do is use the relevant parameters from your
-frontier- model that did converge as starting values for the model
that did not converge. It is upto you or a -frontier- expert on the
list to find out which paremeters are comparable across models, and
thus can be used as starting values. You can set starting values with
the -from()- option, see -help frontier- and -help maximize-.
Alternatively, you can try playing around with the -difficult- or
-technique()- options, and see if your model converges after that.
-- Maarten
-----------------------------------------
Maarten L. Buis
Department of Social Research Methodology
Vrije Universiteit Amsterdam
Boelelaan 1081
1081 HV Amsterdam
The Netherlands
visiting address:
Buitenveldertselaan 3 (Metropolitan), room Z434
+31 20 5986715
http://home.fsw.vu.nl/m.buis/
-----------------------------------------
__________________________________________________________
Sent from Yahoo! Mail.
A Smarter Email http://uk.docs.yahoo.com/nowyoucan.html
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
