Clear message

Basic Arithmetic

[:days4/projects/: Other SAGE Days 4 Project]

• Nick Alexander, Craig Citro, William Stein, David Harvey, David Roe, Joel Mohler, Dave Saunders, Gonzalo Tornaria, Ifti B., Robert Bradshaw
• (Craig Citro) wrap zzp_x from ntl
• (Joel Mohler, David Harvey) improve the ntl wrapper.

• Commutative Diagram Coercion Model: The current SAGE coercion model has some advantages (speed, ability to override for new classes) but also some issues. A user is not able to override the coercion system at runtime and use their own coercion maps (this could be incredibly useful for polynomials, or systems of field extensions). It is difficult to get a system-wide picture of what coercion maps exist: as SAGE grows, this will introduce bugs when the rules for transitivity of coercion are not followed. It is also difficult, though not impossible, to implement coercion systems where the computation of the coercion map requires effort (creating a coercion map from F78→F724). I, David Roe, want to discuss a new coercion model, either replacing over built on top of the old, that addresses these problems. Currently I have some ideas for a system based on commutative diagrams, but I want to brainstorm some more before implementing it. This project will include discussions on ways to improve category theory in SAGE. Nick Alexander is also very interested in working on the proposed commutative diagram coercion model.

Test cases Read these as "an element of S and an element of R"

1. Z[x]/Z∈Q[x] (not Frac(Z[x]))

2. Q+Z[x]∈Q[x] and Z/5Z+Z[x]∈Z/5Z[x]

3. Q*Z[x]∈Q[x]

4. Q*Z/5Z error due to no morphism from all of Q into Z/5Z.

5. Z[x]+Z[y] error due to unknown relation between x and y and ambiguous order

6. Q[ζm]+Q[ζn]∈Q[ζlcm(m,n)] as cyclotomic fields are created with an embedding into Qˉ

7. Fpn+Fpm works using Conway polynomials

8. Z[x]+Q[y] same as 5 (minus the symmetry concerns)

9. Z/16Z+Z/12Z∈Z/4Z

Finding known morphisms

Every parent will hava a list of other objects that it knows how to coerce from (as in the current coerce_try) and known embeddings (which it may append to the lists of existing rings), both created at ring creation time. When trying to find a map from (distinct) R to S we

Creating new morphisms

One can view most desired natural coersions as functorial operations from some simpler base object. E.g.

• Projection/quotients
• Fraction fields/localizations
• Coefficient rings (i.e. polynomial rings, matrix spaces, power series...)
• Forgetful functors
• Field extensions and completions

Given two objects S and R, first see if there is a natural morphism S→R or R→S. If there is one, use it. If there are two, by definition they are cannonically isomorphic and so we can use either.

Considering functors above, decompose each object R as a tower of simpler rings R^0 \rightarrow ···→R′′→R′→R. Now try to complete the pushout diagram R←Y→R for maximal Y as low as possible in the tower of R and S. If the pushout exists, do arithmatic in this ring.

Other

• W Stein: Implement a global proof = True or proof = False option that is queried by algorithms when proof=None.

• (not for this wee) Optimized Polynomial Arithmetic: David Harvey and Bill Hart (Postdoc, Warwick) are in the process of writing FLINT, which features extremely fast univariate polynomial arithmetic over Z. Explore how to fill in the remaining gaps in order to make the results of their work easily available to a larger user community. Similarly, Tom Boothby has worked over the past few months on algorithms for very fast evaluation of polynomials at points, and another project will be to explore how to deploy this.