I am currently working with Jan Brunier and Ken Ono on a practical algorithm to compute the "partition class polynomials" H_n(X) they define in their paper http://arxiv.org/pdf/1104.1182v1, which have trace (24n-1)*p(n). By using using a purely algebraic approach that allows us to work modulo primes that split completely in the ring class field (the so-called CRT method), we hope to avoid the need for any floating point approximations.
It remains to be seen how fast this will be, but I note that this technique has worked well with many other types of class polynomials. Drew Andrew V. Sutherland Research Scientist Department of Mathematics Massachusetts Institute of Technology On Apr 15, 8:55 am, kcrisman <kcris...@gmail.com> wrote: > Just curious if anyone was working on > implementinghttp://aimath.org/news/partition/brunier-onofor > number_of_partitions, > or if this is something useful to do (seems to involve both heavy > modular form work and numeric approximation to ensure algebraic > numbers are sufficiently approximated. > > I'm not suggesting there's anything wrong with "Jonathan Bober's > highly optimized implementation (this is the fastest code in the world > for this problem)." as in the docs, just curious, as these papers > attracted a fair amount of attention. Maybe it's not realistically > even implementable at this time? > > - kcrisman -- To post to this group, send an email to sage-devel@googlegroups.com To unsubscribe from this group, send an email to sage-devel+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/sage-devel URL: http://www.sagemath.org
