|
Size: 4382
Comment:
|
Size: 4384
Comment:
|
| Deletions are marked like this. | Additions are marked like this. |
| Line 5: | Line 5: |
| -- Note: Items 1,2, and 4 are all *broken* until we resolve the bug with simon_two_descent. That is my project now since computing ranks (and descending to \mathbb{Q}: thanks Erin!) will be important in the Elkies-Watkins search over $\mathbb{Q}[i]$. (Jeremy) | -- Note: Items 1,2, and 4 are all *broken* until we resolve the bug with simon_two_descent. That is my project now since computing ranks (and descending to $\mathbb{Q}$: thanks Erin!) will be important in the Elkies-Watkins search over $\mathbb{Q}[i]$. (Jeremy) |
To-do list for elliptic curves in Sage
This is a list of things (small or large) which came up during SD22 at MSRI as deserving to be fixed or implemented in Sage.
-- Note: Items 1,2, and 4 are all *broken* until we resolve the bug with simon_two_descent. That is my project now since computing ranks (and descending to \mathbb{Q}: thanks Erin!) will be important in the Elkies-Watkins search over \mathbb{Q}[i]. (Jeremy)
Regulators over number fields. NB over \mathbf{Q} there are two functions, regulator() and regulator_of_points(), and it is the second of these which can be trivially implemented. Just copy the code from ell_rational_field.py into ell_number_field.py; also copy height_pairing_matrix(). In each case I mean "move" rather than "copy". Then in each case add doctests with a couple of examples over number fields. The other function (regulator) is something which depends on having an actual MW basis, so is not for now. See ticket #9372: JEC has uploaded a patch which is ready for review. (Aly and Jeremy -- reviewed, assuming just the doctest needed changed, positive review)
Linear dependencies of points over number fields (and over Q), modulo torsion. The attached file mwnf.m has Magma code for this (written by me) which may be useful, though you will have to find the Sage equivalent of LLLgram(). That function takes a gram matrix (of not-necessarily full rank) and returns a reduced gram matrix and a matrix in GL(n,Z) and the rank r. (Aly, Jeremy, Jim) \mathbf{Note:} the Sage equivalent of LLLgram() is LLL_gram() which is a method for integer matrices. Second Note: Pari supports LLL_gram for real matrices as well, which we apparently need. There is already a wrapper, but we should add a member method to real matrices (I don't know exactly where this should go yet, haven't looked -- Jeremy).
- Finish ticket #8829 on saturation over number fields. [This depends on item 1.] Probably a job for JEC since he reviewed robertwb's patch and suggested changes, which he will now implement.
- Kilford's rank function patch #9342: completed needs review. (Aly and Jeremy)(Done, has a positive review)
- Integral points over number fields: ongoing project work re-implementing Nook's Magma code.
S-integral points over number fields: ongoing project work, based on Smart-Stephens and taking as a model the code over \mathbb{Q} for S-integral points.
S-class groups: on going project work.
K-Selmer groups esp. K(S,4) and K(S,2)_4: ongoing project work.
- All curves with e.g.r.: putting together a lot of the above! When the time comes, I have some complete lists over imaginary quadratic fields of class number 1 which will be helpful.
All curves with e.g.r. outside S (using most of the above again)
- Iterating through number field elements by height; special case for integers; simpler version not by height. Ongoing project work.
- Heights on projective space over number fields (Jackie's code: make into a ticket+patch)
- Elkies-Watkins over number fields -- post what code you have by the end of the week.
- -- This works very well now. I am working to get access to sage.math to run it on that server (rather than on my computer at home). I need to find a good place, if there is one, to add this into the sage library. Maybe in a miscellaneous file somewhere. (Jeremy)
Implement a 2-torsion rank function over number fields. Jamie is working on this: #9371.
Given an elliptic curves defined over K, determine whether it is isomorphic over K to a curves defined over \mathbb{Q}. [Get the j_invariant, decide if it is rational, if so construct a curve over Q with that j-invariant, base change itback to K, and test for the quadratic twist. All these steps are implemented already. See attached example eg.sage ] (Erin - patch posted #9384)
- I've fixed this function, descend_to. It now returns the correct twist for the general case and works when j=0,1728.
- However, I haven't posted a replacement patch yet because I want the output to be an elliptic curve that sage recognizes is defined over the subfield K of L. This will hopefully be fixed by the end of the day.