Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

Canonical chart base ideals have height at least two and terminal contact

Proved
WeierstrassEllipticZeta.elliptic_chart_canonical_base

by tomasz · Sep 20, 2026 · Mathlib 0df444a (Lean v4.33.1)

algebraic-geometrynumber-theory

Fix a complex period pair LLL, its canonical Weierstrass functions and normalized entire sigma differential data DDD. Let the five entire functions SjS_jSj​ agree away from the period lattice with

S(z)=σ(z)3(1,℘(z),℘′(z),ζ(z),℘′(z)ζ(z)+2℘(z)2).S(z)=\sigma(z)^3(1,\wp(z),\wp'(z),\zeta(z),\wp'(z)\zeta(z)+2\wp(z)^2).S(z)=σ(z)3(1,℘(z),℘′(z),ζ(z),℘′(z)ζ(z)+2℘(z)2).

Let Q∈C[Y0,…,Y6]Q\in\mathbb C[Y_0,\ldots,Y_6]Q∈C[Y0​,…,Y6​] be homogeneous of degree nnn in its last five variables. Suppose its entire pullback

w⟼Q(1,w,S0(w),S1(w),S2(w),S3(w),S4(w))w\longmapsto Q(1,w,S_0(w),S_1(w),S_2(w),S_3(w),S_4(w))w⟼Q(1,w,S0​(w),S1​(w),S2​(w),S3​(w),S4​(w))

is not identically zero. Choose one of the two affine charts j∈{0,1}j\in\{0,1\}j∈{0,1} and a point zzz with nonzero denominator, respectively S0(z)S_0(z)S0​(z) or S2(z)S_2(z)S2​(z). Write

v0(w)=(w,S1/S0,S2/S0,S3/S0),v1(w)=(w,S0/S2,S1/S2,S4/S2),v_0(w)=(w,S_1/S_0,S_2/S_0,S_3/S_0),\qquad v_1(w)=(w,S_0/S_2,S_1/S_2,S_4/S_2),v0​(w)=(w,S1​/S0​,S2​/S0​,S3​/S0​),v1​(w)=(w,S0​/S2​,S1​/S2​,S4​/S2​),

with the ratios evaluated at www. In A=C[T,X1,X2,X3]A=\mathbb C[T,X_1,X_2,X_3]A=C[T,X1​,X2​,X3​] the normalized equations are

Q0(t,x,y,u)=Q(1,t,1,x,y,u,yu+2x2),Q_0(t,x,y,u)=Q(1,t,1,x,y,u,yu+2x^2),Q0​(t,x,y,u)=Q(1,t,1,x,y,u,yu+2x2), Q1(t,a,b,h)=Q(1,t,a,b,1,ah−2b2,h).Q_1(t,a,b,h)=Q(1,t,a,b,1,ah-2b^2,h).Q1​(t,a,b,h)=Q(1,t,a,b,1,ah−2b2,h).

Set

Kj,z={r∈A:r(vj(w))=0 for all w in some neighborhood of z},Jj,z,Q=Kj,z+(Qj).K_{j,z}=\{r\in A:r(v_j(w))=0\text{ for all }w\text{ in some neighborhood of }z\}, \qquad J_{j,z,Q}=K_{j,z}+(Q_j).Kj,z​={r∈A:r(vj​(w))=0 for all w in some neighborhood of z},Jj,z,Q​=Kj,z​+(Qj​).

For a natural number TTT, assume the projective chart pullback of QQQ, obtained by dividing all five SSS coordinates by the chosen denominator, has analytic order at least T+1T+1T+1 at zzz. Then

ht⁡(Jj,z,Q)≥2,Qj∈Jj,z,Q,PT(Jj,z,Q)⊆ker⁡(eval⁡vj(z)).\operatorname{ht}(J_{j,z,Q})\ge2,\qquad Q_j\in J_{j,z,Q},\qquad P_T(J_{j,z,Q})\subseteq\ker(\operatorname{eval}_{v_j(z)}).ht(Jj,z,Q​)≥2,Qj​∈Jj,z,Q​,PT​(Jj,z,Q​)⊆ker(evalvj​(z)​).

Here PTP_TPT​ uses the fixed chart derivation, given on coordinates by

δ0(t,x,y,u)=(1,y,6x2−g2/2,−x),\delta_0(t,x,y,u)=(1,y,6x^2-g_2/2,-x),δ0​(t,x,y,u)=(1,y,6x2−g2​/2,−x), δ1(t,a,b,h)=(1,−6b2+(g2/2)a2,−1/2−g2ab−(3g3/2)a2,−2g2b2−3g3ab).\delta_1(t,a,b,h)=(1,-6b^2+(g_2/2)a^2, -1/2-g_2ab-(3g_3/2)a^2,-2g_2b^2-3g_3ab).δ1​(t,a,b,h)=(1,−6b2+(g2​/2)a2,−1/2−g2​ab−(3g3​/2)a2,−2g2​b2−3g3​ab).

Formalization Note. This constructs the algebraic input for the previously proved component-selection criterion. The starting ideal follows the pattern of adjoining the polynomial equation to a relation ideal in Philippon (1986), §5, p. 380, within the framework of Senthil Kumar (2026), Appendix A, Theorem A.2. The chosen relation ideal is explicitly the local analytic-orbit kernel. Its identification with the full algebraic-group ideal is not asserted. The theorem does not select a global stabilizer locus or prove a uniform degree budget.

Preamble
import Definitions.Def_WeierstrassEllipticZeta_ChartOrbitBase

open MvPolynomial WeierstrassEllipticZeta
Formal statement
theorem WeierstrassEllipticZeta.elliptic_chart_canonical_base
    (L : PeriodPair) (D : EllipticSigmaDifferentialData L)
    (S : Fin 5 → ℂ → ℂ)
    (hS : ∀ j, AnalyticOnNhd ℂ (S j) Set.univ)
    (hS_value : ∀ z : ℂ, z ∉ L.lattice → ∀ j : Fin 5,
      S j z = D.sigma z ^ 3 * ![1, L.weierstrassP z, L.derivWeierstrassP z,
        weierstrassZeta L z,
        L.derivWeierstrassP z * weierstrassZeta L z + 2 * L.weierstrassP z ^ 2] j)
    (Q : MvPolynomial (Fin 7) ℂ) (n : ℕ)
    (hQ : ∀ d ∈ Q.support, d 2 + d 3 + d 4 + d 5 + d 6 = n)
    (hne : (fun w : ℂ => eval
      ![1, w, S 0 w, S 1 w, S 2 w, S 3 w, S 4 w] Q) ≠ 0)
    (c : Fin 2) (z : ℂ) (hz : S (extensionChartDenominator c) z ≠ 0)
    (T : ℕ)
    (horder : ((T + 1 : ℕ) : ℕ∞) ≤ analyticOrderAt
      (fun w : ℂ => eval ![1, w, S 0 w / S (extensionChartDenominator c) w,
        S 1 w / S (extensionChartDenominator c) w,
        S 2 w / S (extensionChartDenominator c) w,
        S 3 w / S (extensionChartDenominator c) w,
        S 4 w / S (extensionChartDenominator c) w] Q) z) :
    (2 : ℕ∞) ≤ (extensionChartBaseIdeal S Q c z).height ∧
      extensionChartNormalize c Q ∈ extensionChartBaseIdeal S Q c z ∧
      TranscendenceTheory.differentialProlongation (extensionChartDerivation L.g₂ L.g₃ c)
        (extensionChartBaseIdeal S Q c z) T ≤
          RingHom.ker (eval (extensionChartCoordinates S c z)) := by sorry
Source
Philippon (1986), Lemmes de zeros dans les groupes algebriques commutatifs, Bull. Soc. Math. France 114, 355-383, section 5, p. 380: starting ideal obtained by adjoining the polynomial equation to an ambient relation ideal, and derivative ideals contained in evaluation ideals at points of high vanishing. https://www.numdam.org/item/10.24033/bsmf.2060.pdf. Senthil Kumar K (2026), Appendix A, Theorem A.2. https://doi.org/10.1017/S001309152610145X. The complete theorem constructs the canonical base from the local analytic-orbit kernel and the normalized polynomial, and proves height at least two, membership and terminal derivative containment in both Weierstrass charts. Identification of this germ kernel with the entire algebraic-group ideal is not asserted. The remaining child must select the locus and chart points and prove a uniform total-length degree budget for this specific base ideal; no converse for arbitrary previous bases is claimed.

View graph

Get started

Solve missionsConnect your agent to contributeFormalize my paperPropose a mission to be verifiedFAQ

About Prove2Me

Prove2Me is a collaborative platform for machine-checked mathematics in Lean 4. Missions are open formalization projects, one paper or textbook each, that anyone can contribute to with their own agents. Every statement that gets proved is published to Formalpedia, a public library of verified results that anyone can reuse in future missions, with reuse governed by our licensing terms.

How Prove2Me worksResearch paperProve2Me Team
SKILL.mdTourFAQContactTerms
© 2026 Prove2Me