Returning to the original question, I suggest that you separate the time taken for the change_ring() operation, which in my experience can take a long time, from the actual computation over ZZ. John
On Monday, 30 March 2026 at 13:24:15 UTC+1 vdelecroix wrote: > Hi Trevor, Dima, > > Interfacing python-flint as we do with cypari2 (over PARI/GP) might be > interesting. It would indeed allow anyone to make a computation > directly in flint via a conversion "sage object" -> "python-flint > object". But it would not solve the problem that the default finite > field sage matrix in the OP has slow multiplication. For that purpose, > it is necessary to have a sage matrix object with a flint backend. > Whether this backend uses flint or python-flint seems secondary to me. > > @Trevor: if you want examples, you could have a look at integer or > rational matrices which are wrappers over the flint types fmpz_mat or > fmpq_mat. > > On Mon, 30 Mar 2026 at 11:34, Georgi Guninski <[email protected]> wrote: > > > > Is the main problem C is significantly faster than python code? > > (I doubt the reason is multiplication with complexity O(n^2.6...)) > > > > -- > > 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 [email protected]. > > To view this discussion visit > https://groups.google.com/d/msgid/sage-devel/CAGUWgD8jAz7r5OgLsvcv_ge0AjKyTCBQ3C_nHweY1RcETNxQCQ%40mail.gmail.com > . > -- 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 [email protected]. To view this discussion visit https://groups.google.com/d/msgid/sage-devel/058fbb51-785b-43cf-9727-25bc2b164ad3n%40googlegroups.com.
