Wrapping up this, for the record, from a couple of weeks back:
On Jun 29, 2007, at 12:48 AM, Robert Bradshaw wrote:
> As other people have mentioned, there's a factor of 2 overhead in the
> first loop. I'm not sure why you're indexing from 1 either.
Thanks for the comments...
Indeed, I managed to inflict a 'typo' while fiddling with various
implementations and systems, so I was indeed incurring a *2 factor in
time. That's fixed.
> As for writing fast pyrex matrix code, M[i,j] is really slow. Here's
> what happens:
Thanks for the details; I went with your version of the code, and
sent William a patch (Trac #400)...
> You will still have the same overhead from object creation for each
> entry (in the integer case), but this would be mitigated by the
> integer pool and should be much faster.
Thanks particularly for Robert's comments on pyrex trickery. Using
Robert's version, it looks like this (now on an Intel Core 2 Duo):
M1 M2 M3
is_symmetric 1.05 s 1.74 s 1.76 s
issym 1.53 s 2.22 s 2.24 s
is_sym3 1.15 s 1.75 s 1.77 s
is_sym4 0.50 s 1.16 s 1.18 s
'is_symmetric' is my original code (with the 'ncols' typo fixed), and
i and j cdef'd (not sure if this was previously the case).
'issym' is "purely pythonic", using xrange().
'is_sym3' is 'issym' in pyrex.
'is_sym4' is Robert's version (this is the version, renamed
is_symmetric, in the patch).
Again, the matrices are symmetric 200x200 matrices with random
elements (so the timings are worst-case :-}).
The limits for the random entries are:
M1: (1,300000)
M2: (1,20000000000000000)
M3: (1,2000000000000000000000000000000000000000000000000000000000000000)
Justin
--
Justin C. Walker, Curmudgeon-At-Large
Institute for the Absorption of Federal Funds
--------
Men are from Earth.
Women are from Earth.
Deal with it.
--------
--~--~---------~--~----~------------~-------~--~----~
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://sage.scipy.org/sage/ and http://modular.math.washington.edu/sage/
-~----------~----~----~----~------~----~------~--~---
