Interesting! I have also noticed that if I just slightly modify the matrix m2, it again works efficiently.
Will someone direct the pariGP guys to this issue? On Thursday, 6 September 2012 17:03:59 UTC+2, D. S. McNeil wrote: > > > That said, I am wondering if this is perhaps a bug in the default > implementation of determinant()? > > It seems strange to me that it takes 8 minutes to compute a determinant > of a 34x34 matrix while other algorithms do it within a second. > > Yeah, it looks like pari's Gauss-Bareiss takes forever on this matrix, > even though its classical Gaussian is quick: > > sage: time m2._det_pari(1) > 336140000000000000 > Time: CPU 0.00 s, Wall: 0.00 s > sage: time m2._det_pari(0) > 336140000000000000 > Time: CPU 597.13 s, Wall: 597.10 s > > Kind of funny: _det_pari(0) is so slow on this particular matrix that > it's five orders of magnitude faster to square m2 and take the root of > the determinant! > > sage: %timeit (m2)._det_pari(1) > 125 loops, best of 3: 3.39 ms per loop > sage: %timeit (m2*m2)._det_pari(0)^(1/2) > 125 loops, best of 3: 5.92 ms per loop > > > Doug > -- You received this message because you are subscribed to the Google Groups "sage-devel" group. To post to this group, send email to sage-devel@googlegroups.com. To unsubscribe from this group, send email to sage-devel+unsubscr...@googlegroups.com. Visit this group at http://groups.google.com/group/sage-devel?hl=en.
