[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]
Re: st: -ml display- behaves differently in Stata 8 & 9
Richard Williams <Richard.A.Williams.5@ND.edu> had trouble getting
exponentiated coefficients with -ml display-.
He also wrote previously about how changes in Stata 9 seemed to break user
written postestimation commands.
> There have been a couple of instances now where user-written programs have
> been "broken" by Stata 9. Most of these seem to be post-estimation
> commands, and hence you have to use version control on the original
> estimation command to avoid the problem. In my case, you would think that
> version control would have avoided the problem, but it didn't; but luckily
> the fix was very simple.
Note that the reason and purpose for version control in Stata is to keep
versioned ado- and do-files from breaking in the future.
In Stata 9, the column names for e(b) and e(V) were changed for -ologit- and
-oprobit-. This change made these commands more consistent with the other
estimation commands in Stata, and eliminated the need to code special cases in
other commands that use the results from these commands (e.g. postestimation
commands, -bootstrap-, -jackknife-, -svy-).
However, versioned do-files written prior to Stata 9 will continue to
work, since the old column names are preserved under version control.
You can do this interactively by calling the command under "version 8.2"
New behavior in -ml display-:
We neglected to do 2 things, leading to Richard's observations about his
1. We didn't mention the new 'eform' behavior nor -e(k_eform)- in the
documentation for -ml display-.
2. Given the nature of the change to -ml display-, we should have thought
better about keeping the old behavior under version control.
Needless to say, we will try to address the above two points before the next
* For searches and help try: