Effective de la Vallée Poussin zero-free region
ProvedZeroFreeBridge.riemannZeta_ne_zero_regionarda-rh65k-transfernumber-theoryriemann-zeta
Let , with , and let be the defined effective constant. Then
This gives the zero-free region without a separate multiplicity assumption.
Preamble
import Definitions.Def_ArdaRH65K_Turing121_DlvpZetaRateEffective
import Definitions.Def_ArdaRH65K_Turing121_StripRepr
import Definitions.Def_ArdaRH65K_Turing121_StripReprR1
import Mathlib
import Theorems.Thm_ZeroFreeBridge_dlvp_zeta_region_rate_effective
import Theorems.Thm_ZeroFreeBridge_zeta_zero_divisor_pos
/- PHASE 4 (dVP frontier, the SELF-CONTAINED zero-free region): drop the multiplicity hypothesis.
`dlvp_zeta_region_rate_effective` takes the zero's multiplicity `k` (with `1 ≤ k`) as input.
`zeta_zero_divisor_pos` derives `1 ≤ divisor ζ (ball c₀ R) ρ₀` directly from `ζ ρ₀ = 0`
(`analyticOrderAt ≥ 1` at a zero, finite because ζ is not locally zero on the connected disk —
`ζ c₀ ≠ 0` at the centre `Re c₀ > 1`). Feeding this in gives the fully self-contained region
`riemannZeta_ne_zero_region` : `55/16 ≤ |γ|` ∧ `1 - dlvpRateC/log|γ| < β` ⟹ `ζ(β+iγ) ≠ 0`,
i.e. `ζ(s) ≠ 0` on `Re s > 1 - dlvpRateC/log|Im s|` (for `|Im s| ≥ 55/16`), with `dlvpRateC` the
EXPLICIT effective constant. The `β ≥ 3/4` restriction of the rate theorem is discharged here:
`dlvpRateC ≤ 1/1792` and `log|γ| ≥ 1`, so `1 - dlvpRateC/log|γ| > 3/4`, hence any zero in the
region already has `β > 3/4`; and `β < 1` since ζ has no zeros with `Re ≥ 1`.
conjecture1_proved = False (NOT a proof of RH — a kernel-verified reduction; it says a zero in the
region cannot exist, which is what dVP gives, not a proof that ζ's zeros lie on the line).
-/
open Complex MeromorphicOn Metric
section
namespace ZeroFreeBridge
end ZeroFreeBridge
open ZeroFreeBridge
Formal statement
theorem ZeroFreeBridge.riemannZeta_ne_zero_region (β γ : ℝ)
(hγ : 55/16 ≤ |γ|) (hβlow : 1 - dlvpRateC / Real.log |γ| < β) :
riemannZeta ((β : ℂ) + (γ : ℂ) * I) ≠ 0 := by sorry
Source