A norm bound for a complex power away from zero
ProvedRSInt.norm_cpow_le_of_gearda-rh65k-transfernumber-theoryriemann-zeta
Let and be real with . For the principal complex power,
This gives a uniform majorant for complex powers on the contour.
Preamble
import Definitions.Def_ArdaRH65K_RS_B2
import Definitions.Def_ArdaRH65K_RS_Entire
import Definitions.Def_ArdaRH65K_RS_Fold
import Definitions.Def_ArdaRH65K_RS_Kernel
import Definitions.Def_ArdaRH65K_RS_Mellin
import Definitions.Def_ArdaRH65K_RS_Strip
import Mathlib
/- RS_B2.lean -- lane RS, ANDURIL brick B2: RIEMANN'S INTEGRAL REPRESENTATION OF zeta(s) for
general complex s, with the residue bookkeeping (the exact Riemann-Siegel formula).
Mathlib-only (imports `RS_Entire`).
## The headline (`riemann_siegel_integral`)
For every complex s that is not an odd positive integer (s ∉ {1, 3, 5, ...}), all N, M ∈ ℕ and
every c1 ∈ (N, N+1), c2 ∈ (M, M+1):
zeta(s) = sum_{n=1}^{N} n^(-s) + chi(s) sum_{m=1}^{M} m^(s-1)
- lineUp c1 (x |-> rsG x x^(-s)) + chi(s) lineDn c2 (x |-> rsH x x^(s-1)),
rsG x = e^{i pi x^2}/(e^{i pi x} - e^{-i pi x}), rsH x = e^{-i pi x^2}/(e^{i pi x} - e^{-i pi x}),
chi(s) = pi^(s-1/2) Gamma((1-s)/2) / Gamma(s/2),
lineUp c F = int F(c + v(1+i))(1+i) dv, lineDn c F = int F(c + v(1-i))(1-i) dv.
This is Riemann's integral (Siegel 1932; Edwards ch. 7; Titchmarsh 2.10) in exactly the form
used for the Riemann-Siegel saddle-point analysis: the remainder integrals run along the
steepest-descent directions of e^{+-i pi x^2} and may cross the real axis anywhere in the gap.
(At s = 3, 5, ... Mathlib's Gamma((1-s)/2) takes the junk value 0 at its pole, so those points
are excluded; s = 1 is the pole of zeta.)
## How it is proved (no residue theorem, no Hankel contour)
1. N = M = 0, Re s > 2 (`rs_identity_re_gt_two`): the Mellin identity (RS_Mellin: Fubini with
the complex Gamma integral, the Mordell closed form, the Bose integral = Gamma zeta) and the
folding of the chi-side line (RS_Fold), assembled with the Gamma identity for chi.
2. All s ∉ {1,3,5,...} (`rs_identity`): identity theorem on the connected open set
ℂ \ {1,3,5,...}; both sides are holomorphic there (RS_Entire).
3. General N, M (`lineUp_book`, `lineDn_book`): induction, crossing one pole at a time
(`lineUp_cross`, `lineDn_cross` in RS_Kernel).
## Critical line (`completedRiemannZeta_eq_rs`, `completed_re_eq_rs_cos`)
For s = 1/2 + it and c ∈ (N, N+1), with S = sum_{n<=N} n^(-s) and R = -lineUp c (rsG x^(-s)):
completedRiemannZeta s = 2 Re( Gammaℝ(s) (S + R) ) (a real number),
and, whenever Gammaℝ(s) = M e^{i phi} with M > 0 (the island's polar form),
Re completedRiemannZeta(s) = 2M ( sum_{n<=N} n^(-1/2) cos(phi - t log n) + Re(e^{i phi} R) ).
So Z(t) = 2 sum n^(-1/2) cos(theta - t log n) + 2 Re(e^{i theta} R_N(t)) EXACTLY: the only
analytic input left for sign certification is a bound on the remainder R_N(t) (brick B3).
conjecture1_proved = False. An exact integral identity for zeta; nothing here bears on RH.
-/
open Complex MeasureTheory Filter Topology Set
open scoped Real ComplexConjugate
noncomputable section
section
namespace RSInt
end RSInt
open RSInt
/-! ## 1. The identity for Re s > 2 -/
/-! ## 2. The domain ℂ \ {1, 3, 5, ...} -/
/-! ## 3. Riemann's integral for every s ∉ {1, 3, 5, ...} (N = M = 0) -/
/-! ## 4. Residue bookkeeping -/
Formal statement
theorem RSInt.norm_cpow_le_of_ge {x w : ℂ} {m : ℝ} (hm : 0 < m) (hx : m ≤ ‖x‖) :
‖x ^ w‖ ≤ ((m ^ w.re + 1) * Real.exp (π * |w.im|)) * Real.exp (|w.re| * ‖x‖) := by sorry
Source