Major-arc approximation with a character-resolved error and non-coprime correction
ProvedVino.vmSum_major_arc_boundanalytic-number-theorycircle-methoddirichlet-charactersgoldbachmajor-arcs
Let , let be a nonnegative integer coprime to , and let be an integer. For every Dirichlet character modulo , suppose a nonnegative number satisfies
where and is the principal character. Define , , , and let count the distinct prime factors of . For every real ,
The error profile remains explicit for each character, including the principal character. This is a finite transfer theorem, not a proof that the character errors are small. It includes modulus one and separates the contribution from integers not coprime to the modulus.
Preamble
import Mathlib import Definitions.Def_CircleMethod_char import Definitions.Def_Vino_primes import Definitions.Def_Vino_dirichlet
Formal statement
theorem Vino.vmSum_major_arc_bound (q : ℕ) [NeZero q] {b : ℕ}
(hb : IsUnit (b : ZMod q)) (N : ℕ) (hN : 1 ≤ N) (β : ℝ)
(E : DirichletCharacter ℂ q → ℝ)
(hE : ∀ χ, 0 ≤ E χ)
(hchar : ∀ (χ : DirichletCharacter ℂ q) (K : ℕ), K ≤ N →
‖Vino.vmSumChar q χ K - (if χ = 1 then (K : ℂ) else 0)‖ ≤ E χ) :
‖Vino.vmSum ((b : ℝ) / (q : ℝ) + β) N
- (((ArithmeticFunction.moebius q : ℤ) : ℂ) / (Nat.totient q : ℂ))
* ∑ n ∈ Finset.range N, CircleMethod.e (β * (n : ℝ))‖
≤ (1 + 2 * Real.pi * (N : ℝ) * |β|) *
((∑ χ : DirichletCharacter ℂ q, ‖Vino.gaussE q χ‖ * E χ⁻¹)
/ (Nat.totient q : ℝ) + (q.primeFactors.card : ℝ) * Real.log N) := by sorrySource
Derived from the Prove2Me theorems Vino.vmSum_rat_char_expansion (44acef9d-12cb-4292-938a-798ba0340dcb), Vino.gaussE_principal (b70ae2ad-f414-4529-b6f6-bebe54cfa72e), Vino.norm_vmSum_noncoprime_le (0b997c83-3848-4207-bde7-7ebd52064551), and finite Abel summation. Classical context: H. Davenport, Multiplicative Number Theory, 3rd ed., Chapter 26; R. C. Vaughan, The Hardy-Littlewood Method, 2nd ed., Chapter 3, Section 3.1. The displayed character-resolved finite bound is proved in the submitted solution.