<>
I am a long-time advocate of do-files to accompany the SJ, so far w/o any success. That would help spot such problems and also make them easy to rectify by just changing the do-file...
HTH
Martin
-----Ursprüngliche Nachricht-----
Von: [email protected] [mailto:[email protected]] Im Auftrag von Nick Cox
Gesendet: Dienstag, 7. Juli 2009 15:29
An: [email protected]
Betreff: RE: st: mvnp code Cappellari & Jenkins 2006
A good way to spot that something was wrong was to notice that `c21' etc are variables, and thus that the code as originally given implied stuffing variables into scalars. That in itself is not illegal in Stata, as Stata would use `c21'[1] etc., but as a general rule that's unlikely to be what you want.
Nick
[email protected]
Maarten buis
--- On Tue, 7/7/09, Charlotte Gary wrote:
> Does anyone have tried the code suggested by Cappellari and
> Jenkins Stata Journal (2006)in paragraph 3.4? (I have
> reported the code at the end)
>
> I am simply trying to run this code (without changes) but
> an error message after “ml maximize” impedes to obtain
> the results that the authors show on p. 170.
You'll need the change the lines:
scalar `cf22' = sqrt( 1 - `c21'^2 )
scalar `cf33' = sqrt( 1 - `c31'^2 - `c32'^2 )
into:
scalar `cf22' = sqrt( 1 - `cf21'^2 )
scalar `cf33' = sqrt( 1 - `cf31'^2 - `cf32'^2 )
and change the line:
egen `sp'=mvnp(`xb1'`xb2'`xb3'), chol(`C') ///
draws($dr) prefix(z) signs(`k1'`k2'`k3')
into:
egen `sp'=mvnp(`xb1' `xb2' `xb3'), chol(`C') ///
draws($dr) prefix(z) signs(`k1' `k2' `k3')
(notice the spaces between variable names)
*
* For searches and help try:
* http://www.stata.com/help.cgi?search
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
*
* For searches and help try:
* http://www.stata.com/help.cgi?search
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
