Fix .multiplication_by_m when m is not coprime to characteristic (#6413)
#37096
Conversation
Compute pth multiplication coordinate maps by using isogenies. Thanks @yyyyx4 for the idea!
|
@yyyyx4 Can you be reviewing this (and other such related PRs from sd123)? |
I wasn't sure if I'm "allowed" to review many of these since I suggested quite a few of them as possible #sd123 projects myself (and in some cases even contributed to the patch with ideas and advice). Is there a policy about this? |
You are definitely allowed to review this. It can even be 2 authors who are the same reviewers but they cross-check each other. Although with the new GH system, this has become more troublesome to do in practice. I can still do a final double-check if you prefer too. |
|
Thanks for the review and sorry for the mess, I didn't really think through the code and just pushed what worked for the first example :P |
| my = ((2*y+a1*x+a3)*mx.derivative(x)/m - a1*mx-a3)/2 | ||
|
|
||
| return mx, my | ||
|
|
||
| def multiplication_by_m_isogeny(self, m): |
There was a problem hiding this comment.
I acknowledge that we are technically "allowed" to remove this method now, but I'm wondering if it's a good idea at this point: Debian and Ubuntu are still on Sage 9.5, so most of their users have never even seen the deprecation warning. Since keeping this method in Sage a little longer is essentially free of harm, I'd suggest to do so.
This reverts commit ac0ce58.
|
I've left the deprecated |
|
Documentation preview for this PR (built with commit 929d0e6; changes) is ready! 🎉 |
| sage: p = 7 | ||
| sage: K.<a> = GF(p^2) | ||
| sage: E = EllipticCurve(K, [K.random_element(), K.random_element()]) | ||
| sage: E.multiplication_by_m(p * 2)[0] == E.multiplication_by_m(p * 2, x_only=True) | ||
| True |
There was a problem hiding this comment.
This test fails at random with probability ~ 1/49 or something like that (i.e. when the discriminant is 0).
There was a problem hiding this comment.
I will fix this in a separate PR since this is merged already. Sorry for not spotting this earlier (there's one similar above)
…ctest
The long-standing issue sagemath#6413 was resolved in sagemath#37096, but there is still
one `# known bug` marker for it in
`src/sage/schemes/elliptic_curves/hom_scalar.py`. Let's remove it.
(Incidentally, the claimed output for the invocation that used to not
work was slightly wrong, too.)
URL: sagemath#37328
Reported by: Lorenz Panny
Reviewer(s):
…ctest
The long-standing issue sagemath#6413 was resolved in sagemath#37096, but there is still
one `# known bug` marker for it in
`src/sage/schemes/elliptic_curves/hom_scalar.py`. Let's remove it.
(Incidentally, the claimed output for the invocation that used to not
work was slightly wrong, too.)
URL: sagemath#37328
Reported by: Lorenz Panny
Reviewer(s):
…ctest
The long-standing issue sagemath#6413 was resolved in sagemath#37096, but there is still
one `# known bug` marker for it in
`src/sage/schemes/elliptic_curves/hom_scalar.py`. Let's remove it.
(Incidentally, the claimed output for the invocation that used to not
work was slightly wrong, too.)
URL: sagemath#37328
Reported by: Lorenz Panny
Reviewer(s):
Almost fixes #6413 - doesn't work for function fields
Compute pth multiplication coordinate maps by using isogenies.
The multivariate substitution step is VERY slow, and I don't know how to fix it. I think the reason might also be the elements live in the fraction field.
Thanks @yyyyx4 for the idea! #sd123 🚀