Hi Jason,
On Thu, Nov 12, 2009 at 04:56:00PM -0500, Jason Bandlow wrote:
> Out of the hospital, but still not able to move around very much,
> unfortunately. It looks like it will be quite a long recovery (4-6
> months or so). But she is in good spirits.
Ouch; that's a long time. My best wishes for her! And bon courage to you!
>> Do you have time to rerun your benchmarks shortly for the SF patch
>> (say today or tomorrow)? If not, may I just set a positive review,
>> assuming you will create a ticket if there remains some performance
>> regression?
>
> Positive review.
Thanks!
> I think there are still some performance issues, but they seem much
> less severe than before. (Kronecker product still seems slow, for
> example.)
Ok. One thing that could have an influence is that the coercion model
currently seems to not always pickup a shortest coercion path (see the
log file on #7420). That should be a fairly easy fix.
> But in the brief time I had to look at this, I was
> getting fairly inconsistent results from trial to trial, and
> multiple trials are slow because I need to restart sage. Actually,
> what I need to do is figure out how to delete the appropriate caches
> from within sage so I can perform many tests in an automated way,
> but I haven't worked that out yet. (Deleting 's' and recreating it
> isn't enough--the cache is very persistent!)
One thing you could try would be to change the base ring. I am not
sure though there is no shared cache (typically, Symmetrica might be
doing some caching). One choice would be QQ[x], QQ[y], ... though
working with polynomials may have a strong influence. Another option
would be a finite but large enough field.
In any case, we need a way to script such benchmark tests; so please
keep notes on how you think it would be best to proceed!
> I hope to have time to look at this in a serious way sometime in the
> next couple of weeks. Then I will create a ticket which is much
> more precise than I am being here.
Thanks. No rush!
Cheers,
Nicolas
--
Nicolas M. Thiéry "Isil" <nthi...@users.sf.net>
http://Nicolas.Thiery.name/
--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups
"sage-combinat-devel" group.
To post to this group, send email to sage-combinat-devel@googlegroups.com
To unsubscribe from this group, send email to
sage-combinat-devel+unsubscr...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/sage-combinat-devel?hl=en
-~----------~----~----~----~------~----~------~--~---
