Question about QUESO's formulation of forward problems

[James Ramsey]

If I look at how Dakota does Monte Carlo sampling for forward problems, it looks like Dakota calculates a bunch of possible input vectors for some computational model. Then the computational model can be run for each vector of inputs (either inside or outside Dakota), possibly in an embarrassingly parallel fashion. Near as I can tell from QUESO's documentation, it looks like a very different and far less parallelizable algorithm is used. Am I reading the documentation correctly, or am I missing something here?

[Paul T. Bauman]

Thanks for your interest in QUESO.

Just to preface, QUESO, at present, is mainly aimed at solving inverse problems. For forward UQ, Dakota has a lot more functionality than QUESO.

On Mon, Jan 30, 2017 at 9:31 AM, James Ramsey <jjra...@pobox.com> wrote:

If I look at how Dakota does Monte Carlo sampling for forward problems, it looks like Dakota calculates a bunch of possible input vectors for some computational model. Then the computational model can be run for each vector of inputs (either inside or outside Dakota), possibly in an embarrassingly parallel fashion. Near as I can tell from QUESO's documentation, it looks like a very different and far less parallelizable algorithm is used. Am I reading the documentation correctly, or am I missing something here?

The interface is the same as for inverse problems, namely, QUESO chooses a point in your domain (in this case via naive Monte Carlo) and passes you that vector. You return the output of your model at that point. QUESO does support parallelism for the forward UQ Monte Carlo - you achieve this by setting the number of sub environments in the QUESO input. That is, QUESO will partition your MPI environment into the number of input sub environments. Then, you can partition that communicator (if you need). So, for example, if you had 100 processors, you could create 100 sub environments and then, assuming your model only needs 1 processor, you could run 100 samples in parallel.

Please let us know if you have more questions.

[Damon McDougall]

Just to add to this, 100 processors and 50 sub environments would mean
each chain owns two processors. This is useful because there are cases
with the likelihood evaluation demands a parallel environment.

> Please let us know if you have more questions.
>

> --
> You received this message because you are subscribed to the Google Groups
> "QUESO-users mailing list" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to queso-users...@googlegroups.com.
> Visit this group at https://groups.google.com/group/queso-users.
> For more options, visit https://groups.google.com/d/optout.


--
Damon McDougall
http://dmcdougall.co.uk
Institute for Computational Engineering Sciences
201 E. 24th St., Stop C0200
The University of Texas at Austin
Austin, TX 78712-1229