Hi,
I am new to sage-devel. Sorry if I asked questions that you perhaps discussed in
the the past.
Why do you wan't to use Strassen for Matrix Multiplication? This algorithms is
not as stable as naive algorithm (if one uses floating point numbers, if you use
it for symbol computations that is no problem). It is also not very fast for
small matrices. And if you don't use the in-memory version of the algorithm it
has extremly memory requirements.
Why not Winograd's method. Or something like Tiling or Loop-Unrolling?
Again, sorry if you discussed this in the past.
Best wishes,
Ivo.
Am 18.07.2011 12:07, schrieb Simon King:
Hi Burcin,
On 18 Jul., 11:37, Burcin Erocal<bur...@erocal.org> wrote:
AFAIK, there are 3 tickets on trac: 4260, 4968, 4258
Thank you!
Perhaps you can help with the pickling, coercion, etc. problems once I
get the linbox wrapper working. ;)
I can at least try.
In the meantime, I work on improving MeatAxe. For example, they don't
use Strassen multiplication. So, it is tempting for me to see whether
MeatAxe can benefit from the stuff in sage.matrix.strassen. For my
application (project on computation of Ext algebras), I will try to
write everything generically, so that eventually both MeatAxe and Sage/
Linbox matrices will work.
Cheers,
Simon
--
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
