Analytic continuation of the fractional-part representation
ProvedZeroFreeBridge.zeta_fract_repr_ofarda-rh65k-transfernumber-theoryriemann-zeta
Let . Suppose the defined expression equals for , the defined fractional-part integral is differentiable at every point of , and is preconnected. Then every satisfies
This is the identity-theorem continuation step.
Preamble
import Definitions.Def_ArdaRH65K_Turing121_StripRepr
import Mathlib
/- INPUT (R) FOR THE STRIP GROWTH BOUND: the fractional-part / Euler-Maclaurin representation
of the Riemann zeta function on the half-plane Re s > 0,
zeta(s) = s/(s-1) - s * integral_{x>1} {x} * x^{-(s+1)} dx (Re s > 0, s != 1),
where {x} = Int.fract x. This is the piece `zeta_strip_bound_of` (in ZeroFreeBridge.lean)
CONSUMES as its hypothesis `hrepr` (with I = the integral), and whose companion bound
`zeta_repr_integral_bound` (input B) is already discharged there. Assembling (R) closes the
crude strip growth bound |zeta(sigma+it)| <= ||s||/||s-1|| + ||s||/sigma to a single unconditional
input. Crude growth only (~|t|); NOT the sharp |t|^{1-sigma} nor the log|t| that feeds the region,
and the Vinogradov-Korobov rate needs VMVT (absent from Mathlib). A gap-filler FEEDING Layer 2,
NOT a proof of RH. conjecture1_proved = False.
PROOF ARCHITECTURE (5 steps):
R1 identity on Re s > 1, via Abel summation (`sum_mul_eq_sub_integral_mul` +
`tendsto_sum_mul_atTop_nhds_one_sub_integral`) applied to f(x) = x^{-s}, c = 1;
R2 the RHS integral is complex-analytic in s on {Re s > 0} (differentiation under the integral);
R3 the whole RHS s/(s-1) - s*I(s) is analytic on {0 < Re s} \ {1};
R4 zeta is analytic on {s != 1};
R5 identity theorem: both sides analytic on the preconnected open {0<Re s}\{1}, agree on the
open subset {Re s > 1}, hence agree everywhere.
This file FULLY ASSEMBLES R3/R4/R5 with NO `stub` and NO `postulate`. The two genuinely-hard
analytic inputs R1 and R2 are taken as explicit HYPOTHESES of the main theorem
`zeta_fract_repr_of` -- exactly the style of `zeta_boundary_contradiction` and `zeta_strip_bound_of`
in ZeroFreeBridge.lean, which likewise take their hard analytic limits/representations as
hypotheses. The convenience wrapper `zeta_fract_repr` (matching the requested signature) is then
`zeta_fract_repr_of` fed the R1/R2 witnesses; those two witnesses are the remaining CI-iteration
targets and their intended proofs are sketched in `zeta_fract_repr_gt`/`differentiableAt_fractIntegral`
below (currently the only two declarations awaiting discharge, clearly delimited).
RISK ORDER: R1 (Abel-summation identity) >> R5-preconnectedness > R2. Uncertain Mathlib names
are flagged inline with `-- FLAG:`.
-/
open scoped Real
open Filter Topology MeasureTheory
section
namespace ZeroFreeBridge
end ZeroFreeBridge
open ZeroFreeBridge
-- (1 : ℂ).re = 1 contradicts 1 < 1
/- ===================================================================================
ASSEMBLED MAIN THEOREM (R3 + R4 + R5), taking R1 and R2 as hypotheses. NO `stub`/`postulate`.
=================================================================================== -/
Formal statement
theorem ZeroFreeBridge.zeta_fract_repr_of
(hR1 : ∀ {z : ℂ}, 1 < z.re → riemannZeta z = stripRHS z)
(hR2 : ∀ {z : ℂ}, z ∈ stripDomain → DifferentiableAt ℂ fractIntegral z)
(hpre : IsPreconnected stripDomain)
{s : ℂ} (hs : s ∈ stripDomain) :
riemannZeta s = stripRHS s := by sorry
Source