So the bulk of the time is spent converting the Pari object back into a Sage object; mainly, this line:
U = self.matrix_space(ncols = self._nrows)([v[0][i,j] for i in xrange(self. _nrows-1,-1,-1) for j in xrange(self._nrows)]) which has 1000x1000 calls to extra the data and then has to be reconstituted into a matrix object. I am not sure if there is a way around this. It might just be a by-product of outsourcing the computation. Best, Travis On Thursday, October 11, 2018 at 5:58:04 AM UTC+10, John Cremona wrote: > > The method smith_form() for matrices over ZZ uses pari's matsnf() but is > vastly slower. Compare these: > > sage: M=MatrixSpace(ZZ,1000,5).random_element() > sage: %timeit S=M.smith_form(transformation=True) > 1 loop, best of 3: 2.57 s per loop > sage: pM=pari(M) > sage: %timeit S=pM.matsnf(1) > 10 loops, best of 3: 49.5 ms per loop > > That's a factor of 50. There's something to be done in the interface > because pari's convention is opposite to Sage's so one has to reverse the > order of rows/columns in the output from pari, but still... > > John > -- You received this message because you are subscribed to the Google Groups "sage-devel" group. To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+unsubscr...@googlegroups.com. To post to this group, send email to sage-devel@googlegroups.com. Visit this group at https://groups.google.com/group/sage-devel. For more options, visit https://groups.google.com/d/optout.
