Hi! Martin Albrecht had opened #9562 and proposed to use M4RIE as default for matrices over GF(2^e). If I understand correctly, the package builds fine on all supported platforms (including Cygwin).
The ticket is almost getting a positive review from me, modulo some minor issues with documentation. On the ticket, David Kirkby reminded us that a new standard spkg requires a vote on sage-devel, so, here we are... I, for one, am particularly impressed by the performance. If I understood correctly, it is not only VASTLY faster than the current implementation in Sage - it even beats Magma. Perhaps Martin can make a stronger case. Here is an example: With the current implementation, we have sage: MS = MatrixSpace(GF(64,'a'),800,800) sage: %time A = MS.random_element() CPU times: user 1.10 s, sys: 0.01 s, total: 1.10 s Wall time: 1.10 s sage: B = MS.random_element() sage: save((A,B),'testmat') sage: %time C=A*B CPU times: user 38.24 s, sys: 0.02 s, total: 38.26 s Wall time: 38.38 s sage: %time A.echelonize() CPU times: user 21.55 s, sys: 0.00 s, total: 21.56 s Wall time: 21.62 s With M4RIE, we have sage: MS = MatrixSpace(GF(64,'a'),800,800) sage: %time A = MS.random_element() CPU times: user 0.02 s, sys: 0.00 s, total: 0.02 s Wall time: 0.02 s sage: B = MS.random_element() sage: %time C=A*B CPU times: user 0.12 s, sys: 0.00 s, total: 0.12 s Wall time: 0.12 s sage: %time A.echelonize() CPU times: user 0.12 s, sys: 0.00 s, total: 0.12 s Wall time: 0.13 s I will certainly use M4RIE in my applications (where I need fast matrices over finite not necessarily prime fields), and I vote for including it as a standard package. Best regards, 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
