Bookmark and Share

Notice: On March 31, it was announced that Statalist is moving from an email list to a forum. The old list will shut down on April 23, and its replacement, statalist.org is already up and running.


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Antwort: Re: RE: Antwort: st: 'mills' output is missing for heckman regression


From   Richard Ochmann <rochmann@diw.de>
To   statalist@hsphsun2.harvard.edu
Subject   Antwort: Re: RE: Antwort: st: 'mills' output is missing for heckman regression
Date   Fri, 14 May 2010 13:52:32 +0200

well, it is all about your first observation as I see it if you call 
-display- here. If the first observation is excluded from your regression 
(which only you know), the Mills ratio will be missing for it. The 
selection probability you however predict out of sample so that it is 
non-missing for your first observation.

best, rich

owner-statalist@hsphsun2.harvard.edu schrieb am 14.05.2010 13:04:31:

> Hi,
> 
> yes that is the problem i am facing with now.
> 
> I don't know why mills200412 is missing. I use the same regression and
> postestimation for another two year data. it works fine. but only this
> one is missing.
> 
> do you know why??
> 
> 
> thanks
> 
> best
> 
> Jing
> 
> On Fri, May 14, 2010 at 6:10 PM, Richard Ochmann <rochmann@diw.de> 
wrote:
> > Nick is right, I was too sloppy here.
> >
> > The output suggests that mills200412[1] is missing, which is why 
Dpr200412
> > [1] turns out to be missing as well, while selpr200412[1] is 
.30568352.
> > I cannot see the reason for this alone from the code given.
> >
> > best, rich
> >
> >> .         predict selpr200412, psel
> >> (32284 missing values generated)
> >>
> >> .         gen Dpr200412=mills200412*(mills200412+selpr200412)
> >> (159277 missing values generated)
> >>
> >> .         display Dpr200412
> >> .
> >>
> >> . display selpr200412
> >> .30568352
> >>
> >> . display mills200412
> >
> >
> > owner-statalist@hsphsun2.harvard.edu schrieb am 14.05.2010 11:54:18:
> >
> >> Not so simple:
> >>
> >> 1. -di varname- is perfectly legal and is equivalent to -di 
varname[1]-.
> >>
> >>
> >> 2. Clearly that is not equivalent to e.g. -list varname-.
> >>
> >> 3. But if a constant were being held in a variable, -di varname- is
> >> sufficient to see what it is.
> >>
> >> 4. That is not especially good in terms of either style or 
efficiency,
> >> but that is another story.
> >>
> >> Nick
> >> n.j.cox@durham.ac.uk
> >>
> >> Richard Ochmann
> >> Sent: 14 May 2010 09:21
> >> To: statalist@hsphsun2.harvard.edu
> >> Subject: Antwort: st: 'mills' output is missing for heckman 
regression
> >>
> >> all I can see here is that at your 'problem 2' you try to -display-
> >> variables which won't work.
> >>
> >> *
> >> *   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/
> >
> 
> 
> 
> -- 
> Best
> 
> 刘京   Jing LIU
> 
> *
> *   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/


© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   Site index