[Yt-dev] Parallelism

Matthew Turk matthewturk at gmail.com
Wed Aug 3 09:59:41 PDT 2011


Hi Stephen,

On Wed, Aug 3, 2011 at 10:41 AM, Stephen Skory <s at skory.us> wrote:
> Matt,
>
>> parallel_concatenate
>> parallel_operation (max, min, etc)
>> parallel_broadcast
>>
>> Would this be a good compromise?
>
> Yes, I think is is a good way to go. It keeps the names clearer
> without being overly specific.

Awesome!  Thanks.  And hopefully we can some day return to the "post
office" idea you and I sketched out a couple years ago.

>
>
>>> I'v already done some of this, and I'd be willing to help out some more.
>>
>> Excellent!  I'd love it if these could show up in a .py file under tests/ .
>
> Didn't I send you that stuff a while ago? Or is that not what you're
> thinking of?

The one you sent me looked like:

http://paste.enzotools.org/show/1735/

Which works great for testing the existing, type-named functionality;
do you think there's any way we could pull back and make it test
integration of the analysis routines, just to make sure that when the
typenames and whatnot are changed, the underlying results remain the
same?  I'll then happily convert those routines I linked above (which
you wrote) to the non-typenamed versions; you were very kind in
writing those, and I am appreciative.  The testing system that is
currently running (through Jenkins on my machine) runs runall.py in
tests, which runs things like these:

http://hg.enzotools.org/yt/src/3fd820235b04/tests/hierarchy_consistency.py
http://hg.enzotools.org/yt/src/3fd820235b04/tests/projections.py
http://hg.enzotools.org/yt/src/3fd820235b04/tests/object_field_values.py

It uses the testing structure we wrote up for Enzo, so it's just
ensuring consistency between versions where there is no actual result
available.

If we could get tests for halo finding and other things that
non-trivially utilize parallelism, that would be pretty much all
that's necessary to start refactoring, I think.

Britton, Cameron: thank you for your thoughts on this.  This could be
a pretty cool expansion of the capabilities.

-Matt

>
> --
> Stephen Skory
> s at skory.us
> http://stephenskory.com/
> 510.621.3687 (google voice)
> _______________________________________________
> Yt-dev mailing list
> Yt-dev at lists.spacepope.org
> http://lists.spacepope.org/listinfo.cgi/yt-dev-spacepope.org
>



More information about the yt-dev mailing list