Addition in complete linear systems of Weil divisors #
This file adds the additive calculus for the complete linear systems defined in
TauCeti.AlgebraicGeometry.WeilDivisor.LinearSystem.Basic.
For an order system S, linear equivalence is compatible with addition of divisors. Hence
members of complete linear systems add:
E ∈ |D|, F ∈ |D'| imply E + F ∈ |D + D'|.
The finite-sum version is the formal divisor bookkeeping used before symmetric powers and Abel maps are available: a finite collection of effective representatives in divisor classes adds to an effective representative in the sum class. The file also records that translating the indexing divisor of a complete linear system by a principal divisor does not change the system.
This advances TauCetiRoadmap/JacobianChallenge/README.md, Layer A, "Divisors on a curve" and
"Degree", by extending the existing abstract complete-linear-system API needed before the
scheme-theoretic symmetric-power and Abel-map layers. No external mathematics is vendored; the
proofs use Tau Ceti's WeilDivisor/OrderSystem API and Mathlib's additive subgroup and
finite-sum lemmas.
Addition of complete linear systems #
Members of complete linear systems add to a member of the complete linear system of the sum class.
Left addition by a fixed member of |D| sends |D'| into |D + D'|.
Right addition by a fixed member of |D'| sends |D| into |D + D'|.
If two complete linear systems are nonempty, then the complete linear system of the sum of their divisor classes is nonempty.
Translating a member of |D| by an effective divisor A gives a member of |D + A|.
Translation by an effective divisor A sends |D| into |D + A|.
Adding an effective divisor to the indexing divisor preserves nonemptiness of complete linear systems.
A finite sum of members of complete linear systems is a member of the complete linear system of the finite sum of the indexing divisors.
Principal translates #
Adding a principal divisor to the indexing divisor does not change the complete linear system.
Subtracting a principal divisor from the indexing divisor does not change the complete linear system.
An effective principal translate of D is a member of the complete linear system |D|.
An effective negative principal translate of D is a member of the complete linear system
|D|.