On Tue, Apr 14, 2009 at 1:15 PM, Ondrej Certik <ond...@certik.cz> wrote: > On Tue, Apr 14, 2009 at 12:43 PM, Fredrik Johansson > <fredrik.johans...@gmail.com> wrote: >> 30 times sounds plausible. Part of the reason is that mpmath doesn't >> use the Euler-Maclaurin formula (I presume PARI does), thus requiring >> more exponentials, and part is the speed of the exponential function. >> The former can obviously be fixed, and the latter could be fixed with >> Mario Pernici's Taylor series code for gmpy. I don't think mpmath can >> be quite as fast as PARI even with these changes, but I think >> the factor could be 3 rather than 30. > > William and Fredrik, whis is your position with regards to gmpy? Do > you want it to ship in Sage?
No, I currently don't want it to ship with Sage though I'm happy with it being an optional spkg (which it is now). > gmpy is quite nice, that it's short, but I hate that it's written in > handwritten C, instead of using Cython. Sage already has Cython > wrappers for gmp, but I don't know how hard would it be to use them > outside of Sage. Currently it would be impossible. > I'd prefer if there could be just one (preferably Cython) wrapper for > gmp and that could be used everywhere. That won't be possible, unfortunately, since Cython doesn't support multiple inheritance. William --~--~---------~--~----~------------~-------~--~----~ 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 For more options, visit this group at http://groups.google.com/group/sage-devel URLs: http://www.sagemath.org -~----------~----~----~----~------~----~------~--~---
