Exact hypersurface-intersection lengths in triangular contact quotients
ProvedWeierstrassEllipticZeta.triangular_hypersurface_intersection_lengthLet , let be an ideal, let be nonzero, and fix three polynomials . Write
Assume the exact membership criterion
For any additional polynomial equation , put and . Then there exists an isomorphism of complex algebras
In particular, is finite-dimensional, with the exact dimension
If , then also
Thus imposing one additional equation on a contact quotient presented by time substitution has an exact length determined by a univariate greatest common divisor.
The substitution map is surjective since it sends to . The hypothesis identifies as the inverse image of the principal ideal . The image of is , and it contains the kernel needed for the quotient correspondence. The first isomorphism theorem produces the displayed algebra isomorphism. Since , its gcd with is nonzero, and the power basis of the univariate quotient gives both finite dimensionality and the degree formula. Divisibility of the gcd into each polynomial gives the two bounds.
Formalization Note No monicity or degree bound on the is required. The result is conditional on the stated ideal-membership criterion, which the elliptic-extension contact presentation supplies. The additional equation may be zero, or its substituted polynomial may be zero; the second bound requires . A nonzero constant is allowed and gives the zero quotient. Lean uses Polynomial.natDegree; the gcd is proved nonzero before finite dimensionality is established, so the dimension formula does not rely on assigning finrank zero to an infinite-dimensional space. This is a local algebraic intersection calculation, not the global zero estimate.
import Mathlib.Algebra.MvPolynomial.Eval import Mathlib.Algebra.Polynomial.AlgebraMap import Mathlib.Data.Complex.Basic import Mathlib.RingTheory.AdjoinRoot import Mathlib.RingTheory.Ideal.Quotient.Operations
theorem WeierstrassEllipticZeta.triangular_hypersurface_intersection_length
(I : Ideal (MvPolynomial (Fin 4) ℂ)) (M : Polynomial ℂ)
(r : Fin 3 → Polynomial ℂ) (hM : M ≠ 0)
(hI : ∀ p : MvPolynomial (Fin 4) ℂ,
p ∈ I ↔ M ∣ MvPolynomial.aeval (Fin.cons Polynomial.X r) p) :
∀ p : MvPolynomial (Fin 4) ℂ,
let q := MvPolynomial.aeval (Fin.cons Polynomial.X r) p
let J := I ⊔ Ideal.span {p}
Nonempty ((MvPolynomial (Fin 4) ℂ ⧸ J) ≃ₐ[ℂ]
(Polynomial ℂ ⧸ Ideal.span {gcd M q})) ∧
FiniteDimensional ℂ (MvPolynomial (Fin 4) ℂ ⧸ J) ∧
Module.finrank ℂ (MvPolynomial (Fin 4) ℂ ⧸ J) = (gcd M q).natDegree ∧
Module.finrank ℂ (MvPolynomial (Fin 4) ℂ ⧸ J) ≤ M.natDegree ∧
(q ≠ 0 → Module.finrank ℂ (MvPolynomial (Fin 4) ℂ ⧸ J) ≤ q.natDegree) := by sorry