On Dec 6, 3:20 pm, Simon King <[EMAIL PROTECTED]> wrote:
> Dear Michael
Hi Simon,
> On Dec 6, 5:55 pm, mabshoff <[EMAIL PROTECTED]> wrote:
>
> > See, this is exactly such a problem that back then was not properly
> > diagnosed and could potentially let someone to conclude that "Sage is
> > crap".
Yes, but back then we either did not boil it down properly or it got
lost in the shuffle. As usual even "Heisenbugs" eventually get tracked
down and hopefully solved one day, so I am very much in favor of
getting them into trac.
> Well, I *did* report it more than one year ago, but I had no trac
> account at that time.
:)
> This time I made a ticket, namely #4731. It provides a short piece of
> code such that the following behaviour occurs:
>
> sage: M=Foo(97,[1,13,100,23098])
> sage: timeit('N=M*13')
<SNIP>
> time for one and the same computation drops by a factor of
> 2 if it is repeated 9 times.
Yes, that should be the same thing.
> The code calls the logarithm (hence, maxima) in a single line. Replace
> that line and the time remains constant.
>
> Good luck with bug-hunting...!
> Cheers
> Simon
I will introduce a new component "performance" in trac shortly since
that fits better than "calculus" in this case since it is less about
correctness than performance. Once we have the needed -timeit
comparison infrastructure in place we will very likely need it more
often :)
Cheers,
Michael
--~--~---------~--~----~------------~-------~--~----~
To post to this group, send email to sage-devel@googlegroups.com
To unsubscribe from this group, send email to [EMAIL PROTECTED]
For more options, visit this group at http://groups.google.com/group/sage-devel
URLs: http://www.sagemath.org
-~----------~----~----~----~------~----~------~--~---
