One step of the unrolled Euler–Maclaurin tail
ProvedZetaReflection.EMHigh.emTail_unroll_steparda-rh65k-transfernumber-theoryriemann-zeta
Let , , and with . Write , let be Bernoulli numbers, and let be the defined rising factorial. Then
This extracts one correction term from the remainder.
Preamble
import Definitions.Def_ArdaRH65K_Turing121_EMZeta
import Definitions.Def_ArdaRH65K_Turing121_EMZetaComplex
import Definitions.Def_ArdaRH65K_Turing121_EMZetaHigh
import Definitions.Def_ArdaRH65K_Turing121_EMZetaTail
import Mathlib.Analysis.Analytic.Uniqueness
import Mathlib.Analysis.Real.Pi.Bounds
import Mathlib.Analysis.SpecialFunctions.ImproperIntegrals
import Mathlib.Analysis.SpecialFunctions.Pow.Deriv
import Mathlib.MeasureTheory.Function.Floor
import Mathlib.MeasureTheory.Integral.IntegralEqImproper
import Mathlib.MeasureTheory.Integral.IntervalIntegral.IntegrationByParts
import Mathlib.NumberTheory.LSeries.RiemannZeta
import Mathlib.NumberTheory.ZetaValues
import Theorems.Thm_ZetaReflection_EMHigh_em_tail_step
/- EMZetaHigh.lean -- the GENERAL-ORDER Euler-Maclaurin representation of ζ with an explicit,
PROVED remainder bound (lane emhigh, memo ANDURIL_ARB_DISCHARGE_2026-09-23 section 3 / step 2
node `AND_em_zeta_orderK`).
The island's order-3 chain (`EMZetaTail.em_zeta_strip3`) stops at the saw-3 remainder. This
file raises the order to ANY `m ≥ 1` by induction, so the finite part carries every Bernoulli
correction `B_2, B_4, …, B_{2K}` and the remainder decays like `N^{-(Re s + 2K - 1)}`:
M1. `poch`, `emC` -- rising factorial `s(s+1)…(s+k-1)` and the k-th x-derivative
coefficient `(-1)^k (s)_k` of `x^{-s}`.
M2. `em_tail_step` -- ONE saw-order-raising step on the tail `[N, ∞)`, any k ≥ 1
(generalises `EMZetaTail.em_tail2_step` / `em_tail3_step`).
M3. `emTail_unroll` -- the induction: the saw-1 tail equals the explicit Bernoulli
sum plus `(-1)^(m-1) emTail m / m!`.
M4. `em_zeta_order` -- the general-order identity for `riemannZeta` on `Re s > 0`.
M5. `emTail_norm_le` -- remainder bound from any sup bound of the order-m saw.
M6. `abs_sawBernoulli_even_le` -- `|B̃_{2K}(x)| ≤ |B_{2K}|` for every K ≥ 1 (Mathlib's
Fourier cosine series of the Bernoulli polynomials).
M6'. `abs_sawBernoulli_odd_le` -- `|B̃_{2K+1}(x)| ≤ 2(2K+1)!/(2π)^{2K+1} (1 + (π²/6−1)/2^{2K−1})`
(Fourier sine series + `ζ(p) ≤ 1 + (π²/6−1)/2^{p−2}`).
M7. `em_zeta_orderK_enclosure` -- THE THEOREM (general K), even-saw remainder:
‖ζ(s) − emFinite K s N‖ ≤ |B_{2K}|/(2K)! · ‖(s)_{2K}‖ · N^{-(σ+2K-1)}/(σ+2K-1),
and `em_zeta_orderK_enclosure_odd`, the same finite part with the saw-(2K+1) remainder
‖ζ(s) − emFinite K s N‖ ≤ 2(1+(π²/6−1)/2^{2K−1})/(2π)^{2K+1} · ‖(s)_{2K+1}‖ · N^{-(σ+2K)}/(σ+2K)
(smaller by about `|s+2K|/(2πN)`), with the evaluator-facing Dirichlet form
`emFinite = Σ_{n<N} n^{-s} + N^{-s}·emCorr`; `em_line_remainder_le` / `_odd_le` are the
critical-line forms with a rational majorant `Q` of the Pochhammer norm and `r ≤ √N`.
M8. exact Bernoulli numbers `B_2 … B_12` (the K ≤ 6 constants).
conjecture1_proved = False. A classical analysis lemma (Euler-Maclaurin), not a proof of RH.
-/
open MeasureTheory intervalIntegral Set Filter Topology Complex
open scoped Real Nat
section
namespace ZetaReflection
end ZetaReflection
open ZetaReflection
section
namespace ZetaReflection.EMHigh
end ZetaReflection.EMHigh
open ZetaReflection.EMHigh
/-! ## M1. Rising factorial and the derivative coefficients of `x^{-s}`. -/
/-! ## M2. The general saw-order-raising step on the tail `[N, ∞)`. -/
/-! ## M3. The induction: unrolling the saw-1 tail to any order. -/
Formal statement
theorem ZetaReflection.EMHigh.emTail_unroll_step {s : ℂ} (hs : 0 < s.re) {N : ℕ} (hN : 1 ≤ N) (j : ℕ) :
(-1) ^ j * emTail s N (j + 1) / ((j + 1)! : ℂ)
= (bernoulli (j + 2) : ℂ) * poch s (j + 1) / ((j + 2)! : ℂ)
* ((N : ℝ) : ℂ) ^ (-s - ((j + 1 : ℕ) : ℂ))
+ (-1) ^ (j + 1) * emTail s N (j + 2) / ((j + 2)! : ℂ) := by sorry
Source