Stata The Stata listserver
[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

RE: st: maximum likelihood procedures for oprobit


From   "David Roodman ([email protected])" <[email protected]>
To   <[email protected]>
Subject   RE: st: maximum likelihood procedures for oprobit
Date   Thu, 8 Dec 2005 21:15:19 -0500

Torgne Middendorf wrote:

I know this question has been adressed before but no one really answered how
to use dummy variables with the xtabond2 command the right way. In xtabond,
there was an option called "diffvars" to prevent Stata from differencing the
concerning variables. In xtabond2 there is a option called "passthru" but
this option only prevents Stata from differencing the instruments (whereas
another question arises, does "passthru" prevent Stata from differencing all
instruments listed after "iv or "gmm"?).

For an answer, see http://www.stata.com/statalist/archive/2005-11/msg00275.html
--David

 -----Original Message-----
From: 	[email protected] [mailto:[email protected]]  On Behalf Of Sunhwa Lee
Sent:	Thursday, December 08, 2005 8:55 PM
To:	Richard Williams; [email protected]
Subject:	Re: st: maximum likelihood procedures for oprobit

Thank you very much for your answer. 

> First off, why do you want to write your own program when Stata 
> already has one?  Even if you get your program to work, Stata's will 
> likely be quicker, more efficient and less likely to encounter problems.

I am working on a more structural model that restricts the way the 
threshold/cutoff points are calculated. So, writing up a program which is 
identical to oprobit, is just the first step to my work. 
 
> Second, while you may have the formulas right, that doesn't mean your 
> code is the same as oprobit's.  The programs go through an iterative 
> procedure, and some procedures will work better than others.

In fact, I am really concerned about this iteration procedures, as the 
estimates on some of my dummy variables are a lot different depending on 
the tolerance levels. 

> Try adding the difficult option to your ml maximize commands, e.g.
> ml maximize, difficult
> According to the help, "difficult specifies that a different stepping 
> algorithm be used in nonconcave regions.  There         is no 
> guarantee that difficult will work better than the default; sometimes 
> it is better, and sometimes it is         worse.  You should use the 
> difficult option only when the default stepper declares convergence 
> and the last         iteration is "not concave", or when the default 
> stepper is repeatedly issuing "not concave" messages and 
> only         producing tiny improvements in the log likelihood."
> 
> You could also try playing around with the technique option.  Type 
> -help maximize- for more details.

I tried "difficult" as well and from "maximize" help, the only two that 
are clearly stated are tolerance and Itorelance levels, which I already 
played around with. 

I found "oprobit.ado" file from stata now and am gonna try to depuzzle 
it. Thank you very much again...

Sunhwa

**************************************
Lee, Sunhwa
PhD Candidate
Department of Economics
University of California at Davis
Davis, CA, 95616
(Office) 109 SSH
(Phone) (530)752-2845
**************************************
*
*   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/

*
*   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/



© Copyright 1996–2024 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index