Lean source theorem: exists_split_primes
ProvedErdosProblems.Shared.QuadraticSplit.exists_split_primeslean-sharedsupport-lemma
In a quadratic extension generated by γ with γ²=b outside the base field, a nonzero modular square root of b at an odd prime ideal of norm ℓ gives two distinct nonzero prime ideals above it, each of the same norm.
Preamble
import Definitions.Def_ErdosProblems_Shared_QuadraticSplitPrimes
import Mathlib.Algebra.CharP.CharAndCard
import Mathlib.Algebra.Field.ZMod
import Mathlib.Algebra.Polynomial.SpecificDegree
import Mathlib.FieldTheory.Minpoly.Field
import Mathlib.NumberTheory.NumberField.Basic
import Mathlib.RingTheory.Ideal.Norm.AbsNorm
import Mathlib.RingTheory.Trace.Basic
/-!
# Degree-one primes that split in a quadratic extension
Let `K ⊆ M` be number fields with `[M : K] = 2` and `M = K(γ)`, `γ ^ 2 = b ∈ 𝓞 K`. A ring
homomorphism `φ : 𝓞 K → ZMod ℓ` (`ℓ` an odd prime) with `φ b = t ^ 2`, `t ≠ 0`, extends to two
distinct ring homomorphisms `𝓞 M → ZMod ℓ`, sending `γ` to `t` and to `-t`.
The ring of integers of `M` is not described explicitly. Instead every `x ∈ M` satisfies
`2 b x = b · Tr(x) + Tr(γ x) · γ` (`Tr = Tr_{M/K}`),
because `x = u + v γ` with `Tr x = 2 u` and `Tr (γ x) = 2 b v`. For `x ∈ 𝓞 M` both traces lie in
`𝓞 K`, so `x ↦ (φ b · φ (Tr x) + φ (Tr (γ x)) · t) / (2 φ b)` is defined on `𝓞 M`; it is
multiplicative because of the two trace identities `trace_identity_one` and
`trace_identity_two`, which follow from the displayed formula.
The kernels of the two extensions are two distinct primes of `𝓞 M` of norm `ℓ`, both lying over
the kernel of `φ`. `exists_split_primes_of_forall_isSquare` packages this in the form consumed by
`ErdosProblems.Shared.IdealCounting.sum_countSupp_mul_le`.
-/
noncomputable section
open NumberField Polynomial
variable {K M : Type*} [Field K] [NumberField K] [Field M] [NumberField M] [Algebra K M]
/-! ### The quadratic extension `M = K(γ)` -/
/-! ### The two extensions of `φ` to `𝓞 M` -/
/-! ### Kernels -/
open ErdosProblems.Shared.QuadraticSplit
Formal statement
theorem ErdosProblems.Shared.QuadraticSplit.exists_split_primes (hfin : Module.finrank K M = 2) (b : 𝓞 K) (γ : 𝓞 M)
(hγ : (γ : M) ^ 2 = algebraMap K M (b : K)) (hγK : (γ : M) ∉ Set.range (algebraMap K M))
{ℓ : ℕ} (hℓ : ℓ.Prime) (hℓ2 : ℓ ≠ 2) (𝔭 : Ideal (𝓞 K)) (φ : 𝓞 K →+* ZMod ℓ)
(hker : RingHom.ker φ = 𝔭) (hnorm : Ideal.absNorm 𝔭 = ℓ)
(t : ZMod ℓ) (ht : t ^ 2 = φ b) (ht0 : t ≠ 0) :
∃ Q₁ Q₂ : Ideal (𝓞 M), Q₁.IsPrime ∧ Q₁ ≠ ⊥ ∧ Q₂.IsPrime ∧ Q₂ ≠ ⊥ ∧
Ideal.absNorm Q₁ = Ideal.absNorm 𝔭 ∧ Ideal.absNorm Q₂ = Ideal.absNorm 𝔭 ∧
Q₁.comap (algebraMap (𝓞 K) (𝓞 M)) = 𝔭 ∧ Q₂.comap (algebraMap (𝓞 K) (𝓞 M)) = 𝔭 ∧
Q₁ ≠ Q₂ := by sorry
Source
Lean source (Apache-2.0): https://github.com/wcook04/plectis-erdos-lean/blob/4fe59e0191169606942baae0949365070b7f419c/ErdosProblems/Shared/QuadraticSplitPrimes.lean#L297-L342
Related paper by Will Cook (CC-BY-4.0): https://github.com/wcook04/plectis-erdos/blob/6917e15ec4abc2623512254da93221e446eeb707/paper/243/erdos-243-reciprocal-tail-rigidity.tex#L1-L58
Paper's AI-assistance disclosure: https://github.com/wcook04/plectis-erdos/blob/6917e15ec4abc2623512254da93221e446eeb707/paper/paper-house-style.sty#L180-L188
Original Erdős problem and Koizumi prior work are distinguished in the paper: https://github.com/wcook04/plectis-erdos/blob/6917e15ec4abc2623512254da93221e446eeb707/paper/243/erdos-243-reciprocal-tail-rigidity.tex#L57-L110