On Mon, Nov 02, 2009 at 03:19:17PM -0500, Jason Bandlow wrote:
> I wanted to give you a quick update on the status of my review. First,
> I haven't been able to look at it as closely as I wanted--Jessica went
> to the hospital of Friday with a blood clot, and that pretty much
> dominated my weekend. (Thankfully, everything seems like it will be
> OK--it was discovered and is being treated before any serious harm was
> done.)
Ouch, that's scary. I cross my fingers that's indeed all fine.
> For what I have looked at so far, it's looking good.
Cool, thanks for looking at it, and thanks for update!
> The main thing that I dislike is the following performance
> degradation. This is in sage-4.1.2 on sage.math, with and without
> the combinat patches applied. Note: I use 'time' instead of
> 'timeit' to measure the speed *without* caching.
>
> ** NEW **
> # Define e,s,h,m,p as SymmetricFunctions(QQ).e(), etc.
> sage: %time res=e(s(h(m(p(s([6,6,1]) + e([7,6]) - p([13]))))))
> CPU times: user 2.30 s, sys: 0.02 s, total: 2.32 s
> Wall time: 2.31 s
>
> ** OLD **
> # Define e,s,h,m,p as SFAElementary(QQ), etc.
> sage: %time res=e(s(h(m(p(s([6,6,1]) + e([7,6]) - p([13]))))))
> CPU times: user 0.20 s, sys: 0.00 s, total: 0.20 s
> Wall time: 0.34 s
>
>
> Do you know if it will be possible to improve this without seriously
> changing the coercion system? If so, this doesn't have to be fixed
> right away, but we should definitely make a ticket. If not, I guess the
> flexibility of the coercion system will be worth it, but it is still
> disappointing.
I doubt the overhead comes from the coercion mechanism: there are very
few coercions in your example. I expect more some stupid thing like
(again!) a mishandled multiplication by scalars somewhere, or modules
morphisms that would need some optimisations. We definitely should put
a ticket for later profiling to find the culprit.
> I'll try to have a formal review for you by tomorrow, as promised.
Thanks!
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
-~----------~----~----~----~------~----~------~--~---
