Dear all,
Thanks for Scott's information and David's method, I solve the initial value
problem in Frontier with the -from()- option .
http://www.stata.com/statalist/archive/2003-05/msg00650.html
However, when I was trying to fit a truncated-normal model, I still have a
problem with option -ufrom()-.
I follow David Drucker's method to feed my own initial values:
1. Run OLS
2. matrix b0 = e(b), ln(e(rmse)^2), .1
3. I add option to the frontier command: distribution(tnormal) ufrom(b0,
copy)
However, I always got a message "ufrom() must specify a matrix".
Is there anything different with -ufrom()- option?
Thanks a lot!
Xun
-----Original Message-----
From: Scott Merryman [mailto:[email protected]]
Sent: Wednesday, July 09, 2003 10:26 PM
To: [email protected]
Subject: st: Re: Initial values not feasible
----- Original Message -----
From: "Wu, Xun" <[email protected]>
To: <[email protected]>
Sent: Wednesday, July 09, 2003 7:50 PM
Subject: st: Initial values not feasible
> Dear all,
>
> When I run Frontier with Stata 8, I got the following message "initial
> values not feasible". I guess this is because ML can't find feasible
initial
> values. I wonder how to solve this problem.
>
> Thanks a lot!
> Brian
>
>
Brian,
This question came up May 28th and David Drukker <[email protected]>
provided a very complete answer. In addition to adjusting the initial
values, make sure you the logs of the variables.
Hope this helps,
Scott
*
* 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/
