L-series
Contents
Project Leader
William
Group Members
William, Cassie, Amy, Lola
Getting Going on Development on *your* laptop (not *.sagenb.org)
Download this file: http://sage.math.washington.edu/home/wstein/days/33/eulerprod.py
Type in the notebook: attach /exact/path/to/eulerprod.py
- Note that elliptic curves over Q(sqrt(5)) won't work if you don't have psage installed.
Project Description
For L-series lovers: Getting the doctest coverage to 100% on this might be a good project:
See this worksheet for an example.
That may discover "issues" (bugs), which I would likely have to fix, but would also be fun because one gets to come up with lots of creative examples of L-series all over the place. Also, the top of that file has a todo list for new features to implement -- most would be bad projects, but one which would be good would be to make it so the Lseries object can use Lcalc (Rubinstein's program) to compute L-series instead of Dokchitser. This would be a good project, because it would mainly involve thinking about the annoying mathematics involved in going between normalizing L-series with the center of the critical strip at 1/2 versus not doing that. Also, it is all pure Python, so easier to get going.
Anyway, I'd say 1 could be a good project for people who know the basics of L-series, but want to get a much more concrete feel for them. In fact, instead of just trying to get coverage to 100%, writing a *tutorial* for computing with L-series using that package would be really nice. E.g., one could walk through how to find missing information, create new L-series classes, etc.
Specific Concrete Projects
- Move eulerprod.py into Sage itself. This would at least involve:
- Clean up the imports at the top of the file.
- Do something with: "from psage.ellcurve.lseries.helper import extend_multiplicatively_generic"; it is the main dependency. It's only a page of code.
- Graceful failure when
- Doctest coverage to 100%
- Fix bugs that are found when doctesting.
- Use rubinstein _lcalc (see line 1116)
Math question: given that you have a_p for Norm(p) <= B, how many bits of precision do we get? Basically, invert the function L.number_of_coefficients(prec=50)
Write an L-functions tutorial: http://sagemath.org/doc/thematic_tutorials This does not have to be long. It has to start somewhere. Small and focused first is fine!
- Speed of elliptic curves over number fields (smalljac).
- Make it so we use exactly one GP session for *all* of the Dokchitser L-functions
- Symmetric powers (and modular degree -- see trac 9758)
- Triple product L-functions: Gross-Kudla, Zhang, etc -- see the code in triple_prod/triple.py
- Make it so we use exactly one GP session for *all* of the Dokchitser L-functions
- Tensor products
- Genus 2 curves, via smalljac and genus2reduction
- Fast L-series of elliptic curves over number fields (not just sqrt(5)), via smalljac
- Inverse of number_of_coefficients function.
Work in progress
L-series tutorial: http://wiki.sagemath.org/days33/lfunction/tutorial
Triple product L-function code: http://wiki.sagemath.org/days33/lfunction/tripleproduct