OAI/NumberTheory/EgyptianFractions/PrimePowerPhaseError.lean

GitHub source file

5 declarations, 0 contain sorry, 0 declared as axioms

  1. theorem OAI.Problem337.norm_mangoldt_sub_prime_phase_sum_le line 12

    Source documentation

    Removing higher prime powers from any bounded complex phase costs at most exactly the elementary Chebyshev error. No distribution of primes is assumed.

    theorem norm_mangoldt_sub_prime_phase_sum_le (N : ℕ) (w : ℕ → ℂ)
        (hw : ∀ n ∈ Finset.Icc 1 N, ‖w n‖ ≤ 1) :
        ‖(∑ n ∈ Finset.Icc 1 N, (ArithmeticFunction.vonMangoldt n : ℂ) * w n) -
          ∑ n ∈ (Finset.Icc 1 N).filter Nat.Prime, (Real.log (n : ℝ) : ℂ) * w n‖ ≤
          Chebyshev.psi N - Chebyshev.theta N := by
    …
  2. theorem OAI.Problem337.primePowerError_mul_log_isLittleO line 30

    Source documentation

    The prime-power discrepancy remains negligible after multiplying by any fixed real power of the logarithm.

    theorem primePowerError_mul_log_isLittleO (D : ℝ) :
        (fun x : ℝ => (Chebyshev.psi x - Chebyshev.theta x) * Real.log x ^ D)
          =o[atTop] (fun x : ℝ => x) := by
    …

    Mentions: OAI.Problem337.D

  3. theorem OAI.Problem337.eventually_norm_mangoldt_sub_prime_phase_sum_le line 44

    Source documentation

    Uniform prime-power deletion with an arbitrary logarithmic saving. The onset is independent of the complex phases.

    theorem eventually_norm_mangoldt_sub_prime_phase_sum_le
        (D ε : ℝ) (hε : 0 < ε) :
        ∀ᶠ N : ℕ in atTop, ∀ w : ℕ → ℂ,
          (∀ n ∈ Finset.Icc 1 N, ‖w n‖ ≤ 1) →
          ‖(∑ n ∈ Finset.Icc 1 N, (ArithmeticFunction.vonMangoldt n : ℂ) * w n) -
            ∑ n ∈ (Finset.Icc 1 N).filter Nat.Prime, (Real.log (n : ℝ) : ℂ) * w n‖ ≤
            ε * (N : ℝ) / Real.log (N : ℝ) ^ D := by
    …

    Mentions: OAI.Problem337.D

  4. theorem OAI.Problem337.norm_prime_phase_sum_le_mangoldt_add_error line 67

    Source documentation

    Any upper bound for the Mangoldt phase sum transfers to the prime-only logarithmically weighted sum with the explicit prime-power error.

    theorem norm_prime_phase_sum_le_mangoldt_add_error (N : ℕ) (w : ℕ → ℂ)
        (hw : ∀ n ∈ Finset.Icc 1 N, ‖w n‖ ≤ 1) :
        ‖∑ n ∈ (Finset.Icc 1 N).filter Nat.Prime, (Real.log (n : ℝ) : ℂ) * w n‖ ≤
          ‖∑ n ∈ Finset.Icc 1 N, (ArithmeticFunction.vonMangoldt n : ℂ) * w n‖ +
            (Chebyshev.psi N - Chebyshev.theta N) := by
    …
  5. theorem OAI.Problem337.prime_phase_log_bound_of_mangoldt line 82

    Source documentation

    Uniform transfer of logarithmic minor-arc estimates. The set of allowed phases may vary arbitrarily with the summation length.

    theorem prime_phase_log_bound_of_mangoldt {ι : Type*}
        (w : ℕ → ι → ℕ → ℂ) (minor : ℕ → Set ι) (D ε : ℝ) (hε : 0 < ε)
        (hw : ∀ N t, t ∈ minor N → ∀ n ∈ Finset.Icc 1 N, ‖w N t n‖ ≤ 1)
        (hbound : ∀ᶠ N : ℕ in atTop, ∀ t ∈ minor N,
          ‖∑ n ∈ Finset.Icc 1 N, (ArithmeticFunction.vonMangoldt n : ℂ) * w N t n‖ ≤
            (ε / 2) * (N : ℝ) / Real.log (N : ℝ) ^ D) :
        ∀ᶠ N : ℕ in atTop, ∀ t ∈ minor N,
          ‖∑ n ∈ (Finset.Icc 1 N).filter Nat.Prime,
            (Real.log (n : ℝ) : ℂ) * w N t n‖ ≤
            ε * (N : ℝ) / Real.log (N : ℝ) ^ D := by
    …

    Mentions: OAI.Problem337.D