Thanks, Robert. I am well advanced with testing this now. I changed 3 files:
1. ~/sage-2.7/data/extcode/pari/simon/ellQ.gp where I changed one function main() to have an extra parameter 2. ~/sage-2.7/devel/sage-main/sage/schemes/elliptic_curves/gp_simon.py which is the associated wrapper, and 3. ~/sage-2.7/devel/sage-main/sage/schemes/elliptic_curves/ell_rational_field.py so that the member function simon_two_descent() for EllipticCurve also has the new parameter. I edited the documentation part of 3 accordingly, and am now trying to find the example which failed before... John On 8/8/07, Robert Bradshaw <[EMAIL PROTECTED]> wrote: > > On Aug 8, 2007, at 4:28 AM, John Cremona wrote: > > > I am using the simon_two_descent() method for EllipticCurve(), and > > have some curves where the default pari precision is in sufficient, > > for example > > > > > > e=EllipticCurve([0,0,0,-10164,409444]); > > e.simon_two_descent() > > > > fails withe the (gp) run-time error " *** bnfsunit: precision too low > > in get_arch." > > > > Now in a normal gp session I can fix this after either > > \p56 > > or > > default(realprecision, 56) > > > > but when running in sage, doing > > pari.set_real_precision(120) > > seems to have no effect. > > > > I guess this is because the latter call effects the precision of the > > pari library functions, whereas simon_two_descent() is a gp script. > > Yes, this is correct. Gp scripts run in their own "clean" session of > pari so that previous commands/etc. won't (adversely or not) affect > their behavior. > > > If this is the case, and there is no better method, I guess we could > > add an extra input parameter "prec" to the simon_two_descent() > > function and then in that function call default(realprecision,prec). > > The default would be 28. > > Sounds good to me. > > > I should be capable of doing this myself! But > > That would be great. > > > (1) I'm not sure which file to edit: looks like > > /home/jec/sage-2.7/devel/sage-main/build/sage/schemes/ > > elliptic_curves/gp_simon.py > > but there is at least one other clone of that file; > > This is the one, where was the other one you were looking at? > > > (2) What should I do then to rebuild sage and test it? > > sage -b rebuilds. To add a test to the doctests, add an example > (preferably one that you fixed) to the area surrounded by triple > quotes at the top of the relevant function. This example will then be > tested before every release of SAGE to make sure it still works > correctly. > > > (3) When I am happy that it works, exactly what should I do to upload > > the change to that it gets incorporated into sage? > > SAGE uses the mercurial revision control system. http:// > www.selenic.com/mercurial/wiki/ When you are happy with your changes, > you can (from the sage prompt) type > > sage: hg_sage.commit() > > which will bring up an editor where you comment on your changes, then > it will check them in. To send those changes to someone else (e.g. > William) do > > sage: hg_sage.bundle(name_of_file) > > and send the resulting .hg file via email. There are lots of other > commands, e.g. hg_sage.diff() to see what you changed, etc. > > This should be enough to get you going, but I would reccoment taking > a look at http://modular.math.washington.edu/sage/doc/html/prog/ > prog.html , specifically section 7. > > > If this is documented, someone please point me in the right direction. > > By the way, whoever it was who did the "SAGE search" thing deserves a > > prize -- it is permanently in my firefox top right corner and is how I > > look up anything about SAGE now! > > > > John > > > > PS I also have some fixes for the mwrank C++ code which I should > > upload, so the answers to my questions above should ideally answer > > that too. > > > > > > -- > > John Cremona > > > > > > > > -- John Cremona --~--~---------~--~----~------------~-------~--~----~ To post to this group, send email to [email protected] 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/ -~----------~----~----~----~------~----~------~--~---
