Hi Simon,
On Tue, Mar 29, 2011 at 12:07:15PM -0700, Simon King wrote:
> Hi!
>
> Tickets #9944 and #9138 provide some nice features, but slow things
> down. It seems to me that the reason of the performance loss is that
> the patches from these tickets make the method resolution order of
> polynomial rings much longer - in some cases the length doubles (15
> versus 39 steps until <type 'object'> is reached).
>
> As much as I understand: If the mro is longer than Python needs more
> time to look up a method that is defined for a very basic class (such
> as sage.structure.category_object.CategoryObject or
> sage.structure.parent.Parent).
I may be wrong but I have the impression that there is a cache here:
Let's create a long mro
class End(object):
def toto(self):
return 1
def long_mro(n):
if n == 0:
return End
else:
class New(long_mro(n-1)):
pass
return New
obj_long = long_mro(500)()
obj_end = End()
Then after that:
sage: len(type(obj_long).mro())
502
sage: len(type(obj_end).mro())
2
sage: timeit('obj_long.toto()',number=10^6)
1000000 loops, best of 3: 187 ns per loop
sage: timeit('obj_end.toto()',number=10^6)
1000000 loops, best of 3: 188 ns per loop
So it seems that there is no difference... Well actually I just figured out
that this has something to do with Cython: if I change
class End(object):
to
class End(SageObject):
Then:
sage: timeit('obj_long.toto()',number=10^6)
1000000 loops, best of 3: 9.22 盜 per loop
sage: timeit('obj_end.toto()',number=10^6)
1000000 loops, best of 3: 203 ns per loop
Now we see the difference. Maybe it's a bug... I'll as on Cython-dev.
Cheers,
Florent
--
To post to this group, send an email to sage-devel@googlegroups.com
To unsubscribe from this group, send an email to
sage-devel+unsubscr...@googlegroups.com
For more options, visit this group at http://groups.google.com/group/sage-devel
URL: http://www.sagemath.org
