Re: st: stcox, shared() takes hours?

[Jeph Herrin <junk@spandrel.net>]

Thanks. I've estimated these models before (with many fewer -id-s)
not noticed them taking any time at all.

cheers,
Jeph


Roberto G. Gutierrez, StataCorp wrote:
> Jeph Herrin <junk@spandrel.net> asks:
>
> > I am estimating (Stata/SE 9.2)
> > . stcox var1 var2, shared(id)
> > for 70k observations on about 1000 -id-s. It has been "Estimating frailty
> > variance:" for approximately 14hrs now. Is that typical? Is there a way to
> > speed this up?  A problem to look for?
> When used with Cox regression, frailties are treated as coefficients to be 
> estimated right along with the betas within a penalized likelihood framework.
> As such, computation time increases with the dimension of the frailties.  With
> 1000 id's, this means your estimation could take a while.  There is neither
> a problem with your Stata nor your syntax.
>
> --Bobby
> rgutierrez@stata.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/
*
*   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/