Explicit polar decomposition of the real gamma factor
ProvedKernelGammaEnvelope.gamma_r_polar_explicitarda-rh65k-transfernumber-theoryriemann-zeta
Let and suppose . Then
This names the magnitude and phase used on the critical line.
Preamble
import Definitions.Def_ArdaRH65K_ThetaConverge
import Definitions.Def_ArdaRH65K_ThetaValue
import Definitions.Def_ArdaRH65K_Turing121_ArctanTaylor
import Definitions.Def_ArdaRH65K_Turing121_KernelGammaEnvelope
import Definitions.Def_ArdaRH65K_XiLineZeros
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.Gamma.BohrMollerup
import Mathlib.Analysis.SpecialFunctions.Gamma.Deligne
import Mathlib.Analysis.SpecialFunctions.Integrals.Basic
import Mathlib.Analysis.SpecialFunctions.Trigonometric.ArctanDeriv
import Mathlib.Analysis.SpecialFunctions.Trigonometric.Bounds
import Mathlib.NumberTheory.LSeries.RiemannZeta
namespace ThetaGap
end ThetaGap
namespace XiLineZeros
end XiLineZeros
/- KernelGammaEnvelope.lean -- ANDURIL G2 kernel discharge: a crude but RIGOROUS two-sided
envelope for the magnitude of the Gamma_R factor on the critical line, and the EXPLICIT
polar decomposition of `gLine` that consumes it.
## Why this file exists
`ZeroSignDecomp_t14.gLine_sign_decomp_t` writes
gLine t = M * (cos phi * Re zeta(1/2+it) - sin phi * Im zeta(1/2+it)), M > 0,
but hides `M` behind an existential, because the sign route never needs its size. A VALUE box
for `gLine t` (the `memR` facts that `CheckBand.checkLine_correct` consumes) does need it: a box
that excludes 0 carries a positive lower bound on `|gLine t|`, hence on `M`. Nothing in the corpus
bounds `|Gamma_R(1/2+it)|` (the `CheckBand` / `ZeroSignDecomp_t14` headers call it "the single
precisely identified blocker"). This file supplies the missing piece in its cheapest honest form:
* `norm_Gamma_le_Gamma_re` : `|Gamma(s)| <= Gamma(Re s)` for `Re s > 0` (Euler's integral and
`norm_integral_le_integral_norm`).
* `norm_sin_le_exp_abs_im` : `|sin z| <= exp |Im z|`.
* `norm_Gamma_quarter_lower` : `(9/4) * exp(-pi |y|) <= |Gamma(1/4 + i y)|`, from the reflection
formula `Gamma(s) Gamma(1-s) = pi / sin(pi s)` and the two upper bounds above.
* `norm_Gamma_quarter_upper` : `|Gamma(1/4 + i y)| <= 4`.
* `gLine_decomp_explicit` : the decomposition with `M = pi^(-1/4) * |Gamma(1/4 + i t/2)|` named.
* `gammaRMag_bounds` : `(3/4) * exp(-pi |t| / 2) <= M <= 4`, `M = gammaRMag t`.
The lower bound is loose (at t = 22 it undershoots the true `M = 3.2e-8` by a factor of about
4e7), but it is a theorem, and it is all a sign-definite box needs: the reflected-band checker
only reads signs.
conjecture1_proved = False. Elementary Gamma-function inequalities, not a proof of RH.
-/
open Complex Real MeasureTheory XiLineZeros ThetaGap
section
namespace KernelGammaEnvelope
end KernelGammaEnvelope
open KernelGammaEnvelope
/-! ## 1. `|Gamma(s)| <= Gamma(Re s)` -/
/-! ## 2. `|sin z| <= exp |Im z|` -/
/-! ## 3. Two-sided bounds for `|Gamma(1/4 + i y)|` -/
/-! ## 4. The explicit polar decomposition of `gLine` -/
Formal statement
theorem KernelGammaEnvelope.gamma_r_polar_explicit (t : ℝ) (Λ : ℝ)
(hbranch : Complex.Gamma (((1/4:ℝ):ℂ) + ((t/2:ℝ):ℂ)*I)
= (‖Complex.Gamma (((1/4:ℝ):ℂ) + ((t/2:ℝ):ℂ)*I)‖ : ℂ) * Complex.exp (Complex.I * (Λ:ℂ))) :
Gammaℝ ((1/2:ℂ) + (t:ℂ)*I)
= ((Real.pi ^ (-(1/4):ℝ) * ‖Complex.Gamma (((1/4:ℝ):ℂ) + ((t/2:ℝ):ℂ)*I)‖ : ℝ) : ℂ)
* Complex.exp (Complex.I * ((Λ - (t/2) * Real.log Real.pi : ℝ):ℂ)) := by sorry
Source