Hi Statalist,
I am running a horse race to compare the estimation of
a Weibull model, a LogLog model and a cox model with
shared frailty. My data has 225 firms. Once a firm
enters the sample, it have the choice of investment or
doing nothing. So this is a multiple failure. After
STSET, the model for parametric hazard runs without
any problem. The Cox model with robust standard errors
adjusted for clustering also runs very fast. But the
Cox model with shared frailty can't converge. I tried
different specifications, taking out one or more
covariates. The message I often encounter is:
numerical derivatives are approximate
flat or discontinuous region encountered
Does anyone know the possible reason and any
suggestions?
Xiao yi
__________________________________________________
Do You Yahoo!?
Tired of spam? Yahoo! Mail has the best spam protection around
http://mail.yahoo.com
*
* 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/
