#117 Random-effects Tobit: turn the attached code into a function package (help wanted)

[Sven Schreiber]

Gretl doesn't have a native facility to estimate a Tobit model on panel data with random effects. (And of course fixed effects are impossible / yield biased estimates.) Jack was kind enough to post a working example implemented in hansl. (See attached file retobit.inp.)
What remains to be done is to package this thing such that it is nicely wrapped and downloadable and useable like the other contributed function packages.
As I said on the list: Microeconometricians to the front!
thanks,
sven

[Sven Schreiber]

There is now a function package "retobit" in the staging area, and hopefully will be approved and then available from within gretl (as all packages are) soon. So I'm closing this.
As noted in the help for the package it is quite slow compared to commercial offerings; we need analytical derivatives of the likelihood to speed things up. Contributors are still welcome even after this ticket is closed.
thanks,
sven