A bound for the shifted digamma remainder
ProvedArgGammaR.abs_remainder_learda-rh65k-transfernumber-theoryriemann-zeta
For , , and real , let and let be the digamma function. Then
This is the error estimate for the gamma-side finite approximation.
Preamble
import Definitions.Def_ArdaRH65K_RSTheta
import Definitions.Def_ArdaRH65K_ThetaConverge
import Definitions.Def_ArdaRH65K_ThetaValue
import Definitions.Def_ArdaRH65K_Turing121_ArctanTaylor
import Definitions.Def_ArdaRH65K_Turing121_ArgGammaR
import Definitions.Def_ArdaRH65K_Turing121_DiffractionCore
import Definitions.Def_ArdaRH65K_Turing121_DlvpTheta
import Definitions.Def_ArdaRH65K_Turing121_DlvpZetaRateEffective
import Definitions.Def_ArdaRH65K_Turing121_RHInBoxAnalytic
import Definitions.Def_ArdaRH65K_Turing121_StripRepr
import Definitions.Def_ArdaRH65K_Turing121_StripReprR1
import Mathlib
import Mathlib.Algebra.Ring.GeomSum
import Mathlib.Analysis.Calculus.Deriv.MeanValue
import Mathlib.Analysis.Complex.ExponentialBounds
import Mathlib.Analysis.PSeries
import Mathlib.Analysis.Real.Pi.Bounds
import Mathlib.Analysis.SpecialFunctions.Complex.Arg
import Mathlib.Analysis.SpecialFunctions.Gamma.Beta
import Mathlib.Analysis.SpecialFunctions.Integrals.Basic
import Mathlib.Analysis.SpecialFunctions.Trigonometric.ArctanDeriv
import Theorems.Thm_ZeroFreeBridge_binetLimit_eq_zero
import Theorems.Thm_ZeroFreeBridge_norm_digamma_sub_log_le_of_anchor
import Theorems.Thm_ZeroFreeBridge_tendsto_binetLimit
namespace RSDesignTheta
end RSDesignTheta
namespace ThetaGap
end ThetaGap
/- ArgGammaR.lean -- brick K6b of the ANDURIL Arb discharge (with the T6 bridge it needs)
(telperion/docs/ANDURIL_ARB_DISCHARGE_2026-09-23.md, inventory row H2f, bricks K6b and T6).
THE BINDERS IT DISCHARGES. Conjuncts 8 and 9 of the `hArbT` binder of every emitted Turing
band (`TuringBand.BandStatement`, `BandGlue.BandData.EnclHyp`) are
hAG1 : DiffractionCore.argChangeVert Gammaℝ (-1) T0 T1 ∈ Set.Icc L4 H4,
hAG2 : DiffractionCore.argChangeVert Gammaℝ 2 T0 T1 ∈ Set.Icc L5 H5,
the argument changes of the archimedean factor `Γℝ(s) = π^{-s/2} Γ(s/2)` up the two vertical
edges of the RvM rectangle. The emitted bands feed them 1e-12-wide Arb enclosures. This file
proves them HYPOTHESIS-FREE for every `0 ≤ T0`, `0 ≤ T1`, with explicit elementary endpoints
(logs and arctans), in the exact binder form:
argChangeVert Γℝ 2 T0 T1 ∈ [S2 T1 - S2 T0 - e2 T1, S2 T1 - S2 T0 + e2 T0] (`hAG2_closed`)
argChangeVert Γℝ (-1) T0 T1 ∈ [Sm1 T1 - Sm1 T0 - em1 T1, Sm1 T1 - Sm1 T0 + em1 T0] (`hAG1_closed`)
S2 T = (1/2) arctan (T/2) + (T/4) log (1 + T²/4) - T/2 - (T/2) log π, e2 T = T/(2(4 + T²)),
Sm1 T = (T/4) log (1/4 + T²/4) + arctan T - T/2 - (T/2) log π, em1 T = T/(2(1 + T²)),
plus the sharper `hAG2_mem` / `hAG1_mem` at any Stirling base index `n0` (error
`y/(4((x+n0)² + y²))` at each edge, `y = T/2`), for the lowest bands.
THE T6 BRIDGE (`integral_half_re_digamma`). `RSTheta.lam_bracket` brackets the GAUSS BRANCH
`Λ(x, y) = lim_n imLnVal x y n` (`gaussLam`), a limit of arctan sums. The binders are integrals
of `logDeriv Γℝ`. The bridge identifies them: for `x > 0` and all `a`, `b`,
∫_a^b (1/2) Re ψ(x + i y/2) dy = Λ(x, b/2) - Λ(x, a/2),
proved WITHOUT any locally-uniform-convergence machinery:
* finite FTC: `d/dy imLnVal x (y/2) n = (1/2)(log n - Σ_{k≤n} (x+k)/((x+k)² + (y/2)²))`;
* `ZeroFreeBridge.digamma_shift`: Re ψ(z) = [that sum] + Re(ψ(z + n + 1) - log n);
* the remainder is at most `(x + 2 + |y|/2)/(2n)` (Binet bound
`ZeroFreeBridge.norm_digamma_sub_log_le` at `Re ≥ 2` plus `log u ≤ u - 1`), so its
integral is `O(1/n)`, and uniqueness of limits closes the identity.
Then `logDeriv Γℝ s = -(log π)/2 + (1/2) ψ(s/2)` gives the σ = 2 edge at `x = 1`, and at σ = -1
the shift `ψ(z) = ψ(z+1) - 1/z` (`z = -1/2 + i y/2`) gives `x = 1/2` plus the elementary
`∫ 1/(1 + y²) = arctan T1 - arctan T0`.
Trust: no hypotheses beyond those stated; axioms [propext, Classical.choice, Quot.sound]
(see AxiomGuardArgChange.lean). No `sorry`.
conjecture1_proved = False. Elementary enclosures of the Γ-factor edges of a finite zero
count; nothing here bears on the Riemann Hypothesis. -/
open Complex Filter Topology
open ThetaGap RSDesignTheta
section
namespace ArgGammaR
end ArgGammaR
open ArgGammaR
/-! ## 1. The Gauss-branch limit `Λ(x, y)` -/
/-! ## 2. The T6 bridge: `∫ (1/2) Re ψ(x + i y/2) dy = Λ(x, b/2) - Λ(x, a/2)` -/
Formal statement
theorem ArgGammaR.abs_remainder_le {x : ℝ} (hx : 0 < x) {n : ℕ} (hn : 1 ≤ n) (y : ℝ) :
|(1 / 2 : ℝ) * ((Complex.digamma ((x : ℂ) + ((y / 2 : ℝ) : ℂ) * I + ((n + 1 : ℕ) : ℂ))).re
- Real.log n)| ≤ (x + 2 + |y| / 2) / (2 * n) := by sorry
Source