Hi,
1000, 10000 parameters? I would go straight to the R and make my own MCMC sampler rather than spending time in making a code for winBUGS. Of course, if you are not very familiar with advanced MCMC methods (like adaptive, HMC, etc.) then it would be hard to make your own sampler in R for such a big problem. Another option would be to use Stan developed by A. Gelman and his posdocs as it is quite powerfull software, which save me several times when my own preprogrammed samplers didn't work as fast as I needed.
However, it's just my very subjective opinion.
Cheers,
Tomas
On 07 Nov 2013, at 18:25, David Ross <[log in to unmask]> wrote:
> I am trying to decide whether or not to spend the time learning BUGS, and the decision hinges on whether or not it would be useful for the problem I am currently working on.
>
> The general question is, roughly how does the computational time with BUGS scale with the number of parameters in a model?
>
> As a start, I have a model that has ~1000 parameters (approximately 85% of which are nuisance parameters). After that I'd probably like to generalize to a model with ~10000 parameters.
>
> Looking at the example problems, I wasn't able to find any that match this kind of scale.
>
> So, accepting that it probably depends on the details somewhat, does anyone have a sense of whether or not BUGS would be able to handle this kind of problem on a high end desktop PC in a reasonable amount of time (less than a day or two)?
>
> Thanks-
>
> -------------------------------------------------------------------
> This list is for discussion of modelling issues and the BUGS software.
> For help with crashes and error messages, first mail [log in to unmask]
> To mail the BUGS list, mail to [log in to unmask]
> Before mailing, please check the archive at www.jiscmail.ac.uk/lists/bugs.html
> Please do not mail attachments to the list.
> To leave the BUGS list, send LEAVE BUGS to [log in to unmask]
> If this fails, mail [log in to unmask], NOT the whole list
-------------------------------------------------------------------
This list is for discussion of modelling issues and the BUGS software.
For help with crashes and error messages, first mail [log in to unmask]
To mail the BUGS list, mail to [log in to unmask]
Before mailing, please check the archive at www.jiscmail.ac.uk/lists/bugs.html
Please do not mail attachments to the list.
To leave the BUGS list, send LEAVE BUGS to [log in to unmask]
If this fails, mail [log in to unmask], NOT the whole list
|