Jean-Fran�ois Godin
>
> I have a large dataset (over 200,000 observations) and I�m
> using a random
> effects model with the command GLLAMM. Has you know GLLAMM
> is a special
> subroutine ado-file which can take a very long time to
> estimate a random
> effects model (in my case with adaptive quadrature estimation).
>
> Running this kind of task on a personal computer is out of
> question (I need
> my computer). So I run my tasks on a UNIX server. Even if
> the UNIX server is
> 3 times slower (32 processors of 300 MHz) than a personal
> computer, it is
> the better option that I have. But the UNIX server has
> sometimes to be stop
> and I need this task to be ran for more than a month! Does
> STATA8 provide a
> command to stop in the middle of a computing task to be
> restarted later
> (example, two days later)?
>
> I�ve checked a few commands but didn�t find anyone. Is
> there a command or
> syntax that would allow me to stop the task after a few
> iterations, save the
> information and restart it after a few days without loosing all the
> �calculation� done before by STATA8?
Never heard of anything remotely resembling this.
> STATA for UNIX allows you to submit multiple tasks, but use only one
> processor for one task. Will it be possible, in a next
> version of STATA, to
> use multiprocessor for only one task, This would largely
> decrease the time
> needed for calculating quadrature's estimate. Mac and Dell offer
> multiprocessors desktop!
>
> Will the next version of STATA will be optimize for
> 64-bytes processors?
Stata Corp rarely answer questions like these.
However, be sure that the interest in such features
is registered.
Nick
[email protected]
*
* 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/
