Is there any way to really implement <= and >= inequality
restrictions? What one can do is to set lnf=. in the user-defined
likelihood program whenever the inequality is not satisfied, but then
if the estimate is at or near the boundary, -ml- stepping around it
hits the areas where the likelihood is missing, and complains that it
cannot compute the necessary derivatives.
Consider this a wish list suggestion if you like: make it possible to
inform -ml- of certain inequality restrictions, and make it possible
to hit the boundary with those, informing the user of the irregular
situations in which the usual inference (standard errors and
likelihood ratios) are not valid.
--
Stas Kolenikov
http://stas.kolenikov.name
*
* 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/
