Re: st: Using a Tobit regression with the Heckman correction

["J. Boreham" <jb648@cam.ac.uk>]

Thank you Nick and Tirthankar, that seems to be working perfectly now!



On Dec 20 2011, Tirthankar Chakravarty wrote:

> As Nick pointed out, your code seems to be incomplete.
>
> Other than that you have got the syntax of -craggit- wrong. I am
> assuming that "fee" is your positive continuous variable? If so, "fee"
> and its determinants go into the "second()" option and the binary
> indicator (your "transferred" variable and its determinants) go first.
> This is in contrast to t the syntax of -heckman-.
>
> T
>
> On Tue, Dec 20, 2011 at 11:26 AM, Nick Cox <njcoxstata@gmail.com> wrote:
> > Your syntax appears truncated as the option -second()- lacks a closing
> > right parenthesis.
> >
> > -craggit-, or more likely -ml-, is having difficulty fitting your
> > model. Try a simpler model or helping out the program by suggesting
> > better initial values.
> >
> > Nick
> >
> > On Tue, Dec 20, 2011 at 7:19 PM, J. Boreham <jb648@cam.ac.uk> wrote:
> >
> > > Thanks again for your help, that does appear to be what I'm after. 
> > > However, when I use it, I get the error message:
> > > "could not calculate numerical derivatives
> > > discontinuous region with missing values encountered"
> > >
> > >
> > > The code I used was:
> > >
> > > ******** craggit fee age agesq curr_app curr_goal_d curr_goal_m 
> > > curr_goal_f, second(transferred age agesq curr_app curr_goal_d 
> > > curr_goal_m curr_goal_f ********
> > > Do you have any idea what may have caused this error?
> > *
> > *   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/