On Friday, January 17, 2014 8:32:20 AM UTC+1, Jeroen Demeyer wrote: > > On 2014-01-16 01:51, ref...@uncg.edu <javascript:> wrote: > > Before I begin, I'd like to thank you in advance for any help I get with > > this question. I'm currently working with some heavy computation and > > decided to cython-ize my code. Because of what I'm doing in my math > > research, I will need to make MILLIONS of calls to pretty high level > > complex functions such as gamma, eta, and the list goes on. I've gotten > > stuck with a curiosity concerning the MPComplexNumber and the > > ComplexNumber types in sage. > That's an artifact of the past (Sage is older than mpc). Next week, I > will have a look at implementing Sage complex numbers with mpc by default. > At some point it was stated on Sage's track (have to search, I think it was a william comment) that this is slower than our implementation using two mpfr_t. And it is still for some basic functionalities (don't really remember, but would say mul). (We had a look together with Andreas at the end of last year and confirmed this.) In particular on the MPC side, it seems the threshold or implem (don't really remember) for karatsuba is not that good. And Sage aslo wins because it does less checks about zeros, infinities, nan and so on. By the way this is not mentioned on the MPFR/MPC todo list for the workshop, but might be something to look at.
> > Jeroen. > -- 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 http://groups.google.com/group/sage-devel. For more options, visit https://groups.google.com/groups/opt_out.
