[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]
Re: st: dtobit and tobcm in Stata 9
At 07:14 AM 6/8/2005 -0700, Alejandro Lopez-Feldman wrote:
I bet this is the same issue that came up the other day with the Long and
Freese spost commands. Stata has changed the way some results are
internally stored, e.g. using equation names when it did not use to do
so. I'm not totally sure what that accomplished - other than breaking old
post-estimation routines!- but I suppose there was some reason for it.
I have tried to use dtobit and tobcm in Stata 9 but I get back the warning
[_se] not found
I guess something changed from 8 to 9 in the way in which standard errors
are saved after tobit... I can get it to work if I set the version to 8.2
but I was wondering if the authors of these ado files are on the list and
if they are aware of the problem.
As you say, you can use the version command, although this might cost you
some 9.0 functionality.
If this is going to be a widespread problem, rather than rewriting all
these old routines (which the authors may or may not choose to do) I wonder
how hard it would be to write a routine that reformats e(b) and e(V) from
9.0 to 8.0 format? I suppose it would depend on whether one conversion
routine would work for everything or whether it would have to be customized
for each of several different commands.
It might also be that rewrites are a minor matter - just search for the old
problematic text and replace it with the appropriate new text.
It would be nice if Stata had a FAQ on updating user-written routines for
Stata 9, which discussed the most common problems and offered possible
solutions. Post-estimation routines seem to be the prime candidates for
problems, since the use of version control may not help unless you also use
the version command with the original estimation routine.
Richard Williams, Notre Dame Dept of Sociology
OFFICE: (574)631-6668, (574)631-6463
WWW (personal): http://www.nd.edu/~rwilliam
WWW (department): http://www.nd.edu/~soc
* For searches and help try: