Fractional-part representation where the zeta series converges
ProvedZeroFreeBridge.zeta_repr_R1arda-rh65k-transfernumber-theoryriemann-zeta
For every with , the defined fractional-part expression obeys
This is the seed identity for continuation to the strip.
Preamble
import Definitions.Def_ArdaRH65K_Turing121_StripRepr
import Definitions.Def_ArdaRH65K_Turing121_StripReprR1
import Mathlib
/- (R1) DISCHARGE for `zeta_fract_repr_of` -- THE HARD ONE, WORK IN PROGRESS.
The fractional-part representation on `Re s > 1`:
riemannZeta s = s/(s-1) - s · ∫_{x>1} {x} x^{-(s+1)} dx = stripRHS s.
Mathlib v4.32.0 has NO such representation (confirmed), so it is built by Abel
summation (`tendsto_sum_mul_atTop_nhds_one_sub_integral₀`) applied to
`f x = x^{-s}`, `c 0 = 0`, `c (n+1) = 1`:
∑_{k≤n} k^{-s} → 0 - ∫_{t>1} (-s·t^{-s-1})·⌊t⌋ dt = s∫_{t>1} ⌊t⌋ t^{-s-1} dt,
⌊t⌋ = t - {t}, s∫_{t>1} t^{-s} dt = s/(s-1) ⟹ ζ(s) = s/(s-1) - s∫_{t>1}{t}t^{-s-1}.
STATUS: FULLY ASSEMBLED, CI-green -- no `sorry`. The five analytic Abel-summation
hypotheses (hf_diff, hf_int, h_lim, hg_dom, hg_int) are discharged on `Re s > 1`
via a shared derivative helper `hasDerivAt_fPow` and the partial-sum `sum_cOne`;
part (A) `tendsto_partialSum_zeta` identifies the LHS limit as `riemannZeta s`
(Complex.summable_one_div_nat_cpow + zeta_eq_tsum... + hasSum_nat_add_iff reindex);
part (B) evaluates the RHS limit value = `stripRHS s` (⌊t⌋ = t - {t} split,
`∫_{Ioi 1}(↑t)^{-s} = 1/(s-1)` via integral_Ioi_cpow_of_lt). This discharges the
R1 input of `zeta_fract_repr_of`. conjecture1_proved = False.
-/
open MeasureTheory Filter Topology Set
section
namespace ZeroFreeBridge
end ZeroFreeBridge
open ZeroFreeBridge
Formal statement
theorem ZeroFreeBridge.zeta_repr_R1 {s : ℂ} (hs : 1 < s.re) : riemannZeta s = stripRHS s := by sorry
Source