Corollary 2.3: the zero-degree boundary branch
ProvedPhilipponMultiplicity.corollary_2_3_zero_degree_boundaryAccepted proof-sketch; geometric coordinate-projection transport remains Open. The Lean reduction proves that zero block degrees remove the corresponding polynomial variables, derives a positive-degree obstruction criterion from Theorem 2.1, lifts the sampling grid and final translate, and chooses a uniform constant over all nonempty factor selections. It explicitly handles S=0 and radii below one. The sole Open child is refcoordinate projection of contact and subgroup obstructions.
This is the zero-degree branch of Corollary 2.3. Let be a Philippon base field and let have disjoint factors, with and . There is a constant , depending only on the embedded product, with the following property.
Let be an analytic subgroup, let , let , and let and the entries of be natural numbers. Suppose at least one . Write . For each tuple , let and be, respectively, the minimum analytic codimension of in and the minimum rank of the image of the sampling group in , over algebraic subgroups not containing and with factor codimensions at least .
Suppose a nonzero multihomogeneous polynomial of multidegree has order at least along at every point of and, for every such ,
Then vanishes on an entire translate of :
This isolates the remaining boundary branch after the positive-degree reduction of Corollary 2.3. It retains , redundant analytic parameters, and every tuple of nonnegative multidegrees with at least one zero entry.
Formalization Note The minima and vanishing order are exactly the mission's existing definitions; a natural-number infimum of an empty family is zero, and natural powers use . This is an open specialization of the published target, not a strengthened hypothesis in Corollary 2.3 itself.
The formal statement is unchanged. The remaining child constructs the projected analytic subgroup with equal contact orders and lifts individual algebraic obstruction subgroups with their tangent-codimension and quotient-rank data. It assumes no multiplicity estimate or vanishing conclusion. The reduction does not compare empty-family minima.
import Definitions.Def_PhilipponMultiplicity_Corollaries set_option autoImplicit false open scoped BigOperators
namespace PhilipponMultiplicity
theorem corollary_2_3_zero_degree_boundary
(K : Type*) [NontriviallyNormedField K] (hK : IsPhilipponBaseField K)
(G : EmbeddedGroupProduct K) (hdisjoint : HasDisjointFactors G) :
∃ c : ℝ, 0 < c ∧
∀ (A : AnalyticSubgroup G) (l : ℕ) (γ : Fin l → G.Point)
(S : ℝ), 0 ≤ S →
∀ (T : ℕ) (D : G.FactorIndex → ℕ) (P : G.CoordinateRing),
(∃ i, D i = 0) →
P ≠ 0 → IsMultihomogeneousOfDegree G P D →
(∀ g ∈ samplingGrid γ ((G.dimension : ℝ) * S),
((G.dimension * T + 1 : ℕ) : WithTop ℕ) ≤ vanishingOrder A P g) →
(∀ r : G.FactorIndex → ℕ, (∀ i, r i ≤ (G.factor i).dimension) →
c * (∏ i, (D i : ℝ) ^ r i) ≤
((T + 1 : ℕ) : ℝ) ^ analyticCodimensionMinimum A r *
S ^ samplingRankMinimum A γ r) →
∃ g : G.Point, translate g A.carrier ⊆ zeroLocusOnGroup G P := by sorry
end PhilipponMultiplicity