Richard Williams <[email protected]> asks how to get a user-written
estimation command to accept the replication based vcetypes, -vce(bootstrap)-
and -vce(jackknife)-:
> One of the things that impresses me about Stata is that it is easy to
> write ml programs for it that support all sorts of options (including
> many that I have no idea what they do!) Anyway, one of my programs
> is called oglm (available from SSC), and the following two commands both work:
>
> oglm warm yr89 male, vce(oim)
> oglm warm yr89 male, vce(robust)
>
> Alas, the following does not:
>
> . oglm warm yr89 male, vce(bootstrap)
> vcetype 'bootstrap' not allowed
>
> So, my question is, what do I need to do to tweak my programs to
> support more of the vce options?
> I don't explicitly define vce as an option in my program. Instead,
> the option gets captured by this command:
>
> mlopts ml_options, `options'
>
> Is mlopts letting some vce options through while blocking others? If
> so, do I need to explicitly include the vce otion in my syntax
> statement, and then check to make sure that the options specified are
> legit? If so, is there some nice code I can just cut and paste?
There is an "undocumented" programmers utility that parses the -vce()- option
for the replication based VCE methods. The "undocumentation" for this parse
utility is available within Stata:
. help _vce_parserun
Based on the help file for -oglm-, I believe the following code segment can be
added to the top of the -oglm.ado- file just before the -version- statement.
***** BEGIN:
if _by() {
local BY `"by `_byvars'`_byrc0':"'
}
`BY' _vce_parserun oglm, mark( Exposure ///
OFFset ///
CLuster ///
HETero ///
SCALE ///
eq2) : `0'
if "`s(exit)'" != "" {
version 9.2: ereturn local cmdline `"oglm `0'"'
exit
}
***** END:
--Jeff
[email protected]
*
* 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/
