bug in elliptic curve two_descent command

[williamstein]

Don't do it twice:

sage: E = EllipticCurve([1,1,0,0,528])
sage: E.two_descent(verbose=False)
True
sage: E.two_descent(verbose=False)
---------------------------------------------------------------------------
RuntimeError                              Traceback (most recent call last)

/Users/wstein/<ipython console> in <module>()

/Users/wstein/sage/install/sage-4.6.1/local/lib/python2.6/site-packages/sage/schemes/elliptic_curves/ell_rational_field.pyc in two_descent(self, verbose, selmer_only, first_limit, second_limit, n_aux, second_descent)
    868                         n_aux, second_descent)
    869         if C.certain():
--> 870             self.__gens[True] = [self.point(x, check=True) for x in C.gens()]
    871             self.__gens[True].sort()
    872             self.__rank[True] = len(self.__gens[True])

/Users/wstein/sage/install/sage-4.6.1/local/lib/python2.6/site-packages/sage/libs/mwrank/interface.pyc in gens(self)
    598         from sage.misc.all import preparse
    599         from sage.rings.all import Integer
--> 600         return eval(preparse(self.__two_descent_data().getbasis().replace(":",",")))
    601 
    602     def certain(self):

RuntimeError: 

Component: elliptic curves

Author: Peter Bruin

Branch/Commit: b437f35

Reviewer: Frédéric Chapoton

Issue created by migration from https://trac.sagemath.org/ticket/10665

[sagetrac-weigandt]

comment:1

This is probably explained by Cremona's comment at #10108:

"I think I can explain this. The problem to be solved for this ticket was that if mwrank is given incomplete but otherwise correct input, it just waits for the rest of the input, making Sage appear to hang. To fix this I made sure that the input provided by Sage always has some other stuff appended to it, so mwrank never has insufficient input. But then, the next time input is sent to mwrank, there is likely to be still some of that extra stuff in its input buffer. To get around that (I thought) I made sure that mwrank was restarted at every call.

Clearly what I did was insufficient, but this does explain who the order of executing commands does matter."

[JohnCremona]

comment:2

And there are patches at #10108 which represent a lot of work which have been lying about unused and unmerged just because they are not perfect, even if Sage is better with them than without.

[chriswuthrich]

comment:5

Even after #10108, this is still a problem !

[pjbruin]

comment:7

The cause is that each time you call mwrank_EllipticCurve.two_descent(), it creates a new __descent (the 2-descent data) but does not reset the __saturate bound. Therefore no saturation is done on the new __descent data, which is bad because getbasis() depends on the saturation having been done. I'm now testing a patch.

[pjbruin]

Branch: u/pbruin/10665-mwrank_elliptic_curve_two_descent

[pjbruin]

Author: Peter Bruin

[pjbruin]

comment:8

(Note: the change to if not self.__descent.ok() is just a trivial simplification, the important line is self.__saturate = -2.)

[pjbruin]

Commit: b437f35

[fchapoton]

comment:9

ok, the patchbot is green, and the patch looks simple enough to me.

So let me give a positive review.

[fchapoton]

Reviewer: Frédéric Chapoton

[vbraun]

Changed branch from u/pbruin/10665-mwrank_elliptic_curve_two_descent to b437f35