---------- Forwarded message ---------- From: John Cremona <[EMAIL PROTECTED]> Date: 26 Dec 2007 22:48 Subject: Re: cremona package under Sage To: [EMAIL PROTECTED]
I cannot help with any of that since I did none of the Sage-mwrank interface myself at all, and you have already found out more about how that works than I know. I realize that it would be better if I did know about it all as then I could maintain it and add features myself, instead of (as at present) implementing desired features in my C++ code and waiting for someone else to add them to the interface. I think it is a problem that these interfaces are not well documented (hence my CC-ing the sage-devel list. The convention is not to go off list as long as the discussion involves Sage at all, as it is easy or other subscribers to ignore threads which do not interest them). It's funny what you say about "elliptic curves defined over ZZ rather than QQ". In my C++ code at present only integral models are allowed, which is what I suppose you mean by being "defined over ZZ" (though strictly speaking there are no elliptic curves defined over ZZ -- that's a theorem since such an object would have to have discriminant a unit in Z and there just aren't any!). One feature which I have been asked to add, which I was thinking about before the holiday, is how t extend at least some of the functionality I provide to non-integral models. I wlll get back to that after the holiday, but not for several days since I am tied up with family affairs. What I can easily implement is the C++ code is either of the following: an option flag integral_points_only (default 0), or alternatively a denominator_bound. Currently tha maximum denominator is set in qcurves/sieve_search.cc:89 (b_high=height) which would be instead either b_high=1 or b_high=denominator_bound. Then the Sage interface would need to amended correspondingly. John On 26/12/2007, bill purvis <[EMAIL PROTECTED]> wrote: > John, > you did say to contact you if I have difficulty understanding your code: > > My current problem is tying together the Python code with the > C++ in the cremona package. > > The code for point_search() in ell_rational_field.py contains the > following three lines: > > mwrank = mwrank.mwrank_MordellWeil(c, verbose) > mw.search(height_limit, verbose=verbose) > v = mw.points() > > I tracked down mwrank_MordellWeil to .../sage/libs/mwrank/interface.py > > and found it in turn creates a struct _mw which has .search() defined > as search. This in turn I found in .../sage/libs/mwrank/mwrank.pyx > and this invokes mw_search(). Somewhere this must tie in to the > C++ code, but I've not yet found any definition for it. > Can you speed up the process by giving me a brief outline of how > it all fits together. I had thought in terms of just adding an extra > argument but there are so many layers here that I wonder if that is > appropriate. I suppose that an alternative approach would be to > define the elliptic curve over ZZ rather than QQ, but I notice that > this would require substantial effort (which I might consider later) > as basic things like defining a point on such a curve gives a > 'NotImplementedError'. > > Bill > > -- > +---------------------------------------+ > | Bill Purvis, Amateur Mathematician | > | email: [EMAIL PROTECTED] | > | http://bil.members.beeb.net | > +---------------------------------------+ > -- John Cremona -- John Cremona --~--~---------~--~----~------------~-------~--~----~ To post to this group, send email to sage-devel@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/sage-devel URLs: http://sage.scipy.org/sage/ and http://modular.math.washington.edu/sage/ -~----------~----~----~----~------~----~------~--~---
