Hi Antonio,
I once wrote a generalization of -clad- that uses the same estimation
technique (Buchinsky's iterative linear programming algorithm) but
allows a different censoring point for each observation. The program
is called -cqreg-. I will the send you a copy off-list. Please note
that it is not clear whether the algorithm produces consistent results
in the case of variable censoring points. It has been on my list since
long to investigate on this, but I just never got to do it. This is
why -cqreg- is not public yet. Use it at your own risk (i.e., if I
were you, I would conduct some simulations to evaluate whether -cqreg-
produces appropriate results in your situation).
ben
On 6/10/07, Antonio R Carrizo <[email protected]> wrote:
Dear all,
The CLAD estimator, implemented by Jolliffe et al., is certainly an
interesting alternative to TOBIT. Moreover, the cnreg estimator in
Stata is extremely useful since it allows different censoring points.
I was wondering if anyone has seen or done a CLAD estimator where
different censoring points are allowed (I couldn't find it in my
searches) - i.e., some form of "CNLAD" estimator.
Many thanks
*
* 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/
*
* 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/
