Proof AtlasMCP access
ProofsNumber theory

Lean proof review
The Riemann Hypothesis

Unverified claim

A proposed Lean proof that every nontrivial zero of ζ(s) has real part ½.

Lean 4Number theoryƒDirichlet L-functions
Development/AutomorphicArithmetic/DedekindLFunctions/LandauTheorems/mellinGchiModOne.lean
1import Development.AutomorphicArithmetic.DedekindLFunctions.LandauZeroFreeCriterion2import Development.AutomorphicArithmetic.DedekindLFunctions.LandauTheorems.MellinRamp34open Complex MeasureTheory Set Filter Topology56noncomputable section78namespace Deduction910theorem mellin_Gchi_modOne_sum {μ : Measure ℝ} (T : ℕ → ℝ → ℂ) (E : ℝ → ℂ)11    (hT : ∀ n, Integrable (T n) μ) (hE : Integrable E μ)12    (hs : Summable fun n => ∫ a, ‖T n a‖ ∂μ) (hp : ∀ a, Summable fun n => T n a) :13    ∫ a, (E a + ∑' n, T n a) ∂μ = ∫ a, E a ∂μ + ∑' n, ∫ a, T n a ∂μ := by14  let F : ℕ → ℝ → ℂ := fun n => Nat.casesOn n E T15  have hF : ∀ n, Integrable (F n) μ := fun n => by16    cases n with17    | zero => exact hE18    | succ n => exact hT n19  have hFs : Summable fun n => ∫ a, ‖F n a‖ ∂μ := (summable_nat_add_iff 1).mp hs20  have h := hasSum_integral_of_summable_integral_norm hF hFs21  have h2 : ∀ a, ∑' n, F n a = E a + ∑' n, T n a := fun a =>22    ((summable_nat_add_iff (f := fun n => F n a) 1).mp (hp a)).tsum_eq_zero_add23  simp_rw [h2] at h24  rw [← h.tsum_eq, h.summable.tsum_eq_zero_add]25  rfl2627theorem mellin_Gchi_modOne_int (n : ℕ) (s : ℂ) (hs : 1 < s.re) :28    IntegrableOn (fun y : ℝ => ((max (y / n - 1) 0 : ℝ) : ℂ) * (y : ℂ) ^ (-s - 1)) (Ioi 1) := by29  have ha : (-s).re < -1 := by simp; linarith30  have hb : (-s - 1).re < -1 := by simp; linarith31  refine Integrable.mono' (integrableOn_Ioi_norm_cpow_of_lt ha zero_lt_one) ?_ ?_32  · refine AEStronglyMeasurable.mul ?_ (integrableOn_Ioi_cpow_of_lt hb zero_lt_one).133    exact Continuous.aestronglyMeasurable (by fun_prop)34  · refine ae_restrict_of_forall_mem measurableSet_Ioi fun y hy => ?_35    have hy1 : (1 : ℝ) < y := hy36    have hy0 : (0 : ℝ) < y := by linarith37    have hyC : (y : ℂ) ≠ 0 := by exact_mod_cast hy0.ne'38    have h1 : y / n ≤ y := by39      rcases Nat.eq_zero_or_pos n with rfl | hn40      · simpa using hy0.le41      · exact div_le_self hy0.le (by exact_mod_cast hn)42    have h2 : max (y / n - 1) 0 ≤ y := max_le (by linarith) hy0.le43    rw [norm_mul, cpow_sub _ _ hyC, cpow_one, norm_div, Complex.norm_real, Complex.norm_real,44      Real.norm_of_nonneg (le_max_right _ _), Real.norm_of_nonneg hy0.le]45    calc _ ≤ y * (‖(y : ℂ) ^ (-s)‖ / y) := by gcongr46      _ = _ := by field_simp4748theorem mellin_Gchi_modOne_term (n : ℕ) (s : ℂ) (hs : 1 < s.re) :49    ∫ y in Ioi (1 : ℝ), ((ArithmeticFunction.vonMangoldt n : ℝ) : ℂ) *50      (((max (y / n - 1) 0 : ℝ) : ℂ) * (y : ℂ) ^ (-s - 1)) =
203 linesLn 10UTF-8Raw