OAI/GroupTheory/PolycyclicRecognition/Analysis/PairSublinear.lean

GitHub source file

5 declarations, 0 contain sorry, 0 declared as axioms

Sublinear height errors from the coupled transfers

The finite invariant measure on the inverse section need not be ergodic. The transfers already constructed there suffice for the same word estimate.

  1. theorem OAI.PolycyclicRecognition.isCocycle_pairHeight_sub_hom line 31

    Source documentation

    Subtracting any continuous height homomorphism preserves the cocycle identity, independently of any ergodicity or averaging assertion.

    theorem isCocycle_pairHeight_sub_hom
        (π l : G →* Multiplicative W) (hπ : Continuous fun g => (π g).toAdd) :
        IsCocycle (fun g => pairHeightL2 μ π hπ g - Lp.const 2 μ (l g).toAdd) := by
    …

    Mentions: OAI.PolycyclicRecognition.IsCocycle, OAI.PolycyclicRecognition.pairHeightL2

  2. theorem OAI.PolycyclicRecognition.pairHeight_coboundary_of_squared_errors line 47

    Source documentation

    A compact-uniform squared-error estimate gives an actual small coboundary in the Hilbert space, for a finite invariant measure.

    theorem pairHeight_coboundary_of_squared_errors
        (π l : G →* Multiplicative W) (hπ : Continuous fun g => (π g).toAdd)
        (φ : ℕ → NormalizedPair G Λ → W) (hφ : ∀ n, MemLp (φ n) 2 μ)
        (Q : Set G)
        (happrox : Filter.Tendsto (fun n => ⨆ t ∈ Q,
          ∫⁻ p, ‖pairTransferDefect π l (φ n) t p‖ₑ ^ 2 ∂μ) Filter.atTop (nhds 0)) :
        ∀ ε : ℝ, 0 < ε → ∃ v : Lp W 2 μ, ∀ q ∈ Q,
          ‖(pairHeightL2 μ π hπ q - Lp.const 2 μ (l q).toAdd) + q • v - v‖ ≤ ε := by
    …

    Mentions: OAI.PolycyclicRecognition.NormalizedPair, OAI.PolycyclicRecognition.pairTransferDefect, OAI.PolycyclicRecognition.pairHeightL2

  3. theorem OAI.PolycyclicRecognition.pairHeight_sublinear_of_squared_errors line 85

    Source documentation

    Compact-uniform transfer errors imply the global sublinear word bound. This applies directly to the finite measure constructed by inverse coupling.

    theorem pairHeight_sublinear_of_squared_errors
        (π l : G →* Multiplicative W) (hπ : Continuous fun g => (π g).toAdd)
        (φ : ℕ → NormalizedPair G Λ → W) (hφ : ∀ n, MemLp (φ n) 2 μ)
        (Q : Set G) (length : G → ℝ) (hlen : ∀ g, 0 ≤ length g)
        {a : ℝ} (ha : 0 ≤ a)
        (hword : ∀ g, ∃ w : List G, (∀ q ∈ w, q ∈ Q) ∧ w.prod = g ∧
          (w.length : ℝ) ≤ a * length g + a)
        (happrox : Filter.Tendsto (fun n => ⨆ t ∈ Q,
          ∫⁻ p, ‖pairTransferDefect π l (φ n) t p‖ₑ ^ 2 ∂μ) Filter.atTop (nhds 0)) :
        ∀ ε : ℝ, 0 < ε → ∃ B : ℝ, 0 ≤ B ∧ ∀ g,
          ‖pairHeightL2 μ π hπ g - Lp.const 2 μ (l g).toAdd‖ ≤ ε * length g + B := by
    …

    Mentions: OAI.PolycyclicRecognition.NormalizedPair, OAI.PolycyclicRecognition.pairTransferDefect, OAI.PolycyclicRecognition.pairHeightL2

  4. theorem OAI.PolycyclicRecognition.pairHeight_sublinear_of_minimizing_paths line 101

    Source documentation

    For a proper group with minimizing paths, compact-uniform transfer approximation alone supplies metric sublinearity.

    theorem pairHeight_sublinear_of_minimizing_paths
        (π l : G →* Multiplicative W) (hπ : Continuous fun g => (π g).toAdd)
        (φ : ℕ → NormalizedPair G Λ → W) (hφ : ∀ n, MemLp (φ n) 2 μ)
        (happrox : ∀ Q : Set G, IsCompact Q →
          Filter.Tendsto (fun n => ⨆ t ∈ Q,
            ∫⁻ p, ‖pairTransferDefect π l (φ n) t p‖ₑ ^ 2 ∂μ) Filter.atTop (nhds 0))
        (hpath : ∀ g : G, ∃ γ : ℝ → G, γ 0 = 1 ∧ γ 1 = g ∧
          ∀ t ∈ Set.Icc (0 : ℝ) 1, ∀ u ∈ Set.Icc (0 : ℝ) 1,
            dist (γ t) (γ u) ≤ dist 1 g * |t - u|) :
        ∀ ε : ℝ, 0 < ε → ∃ B : ℝ, 0 ≤ B ∧ ∀ g,
          ‖pairHeightL2 μ π hπ g - Lp.const 2 μ (l g).toAdd‖ ≤ ε * dist 1 g + B := by
    …

    Mentions: OAI.PolycyclicRecognition.NormalizedPair, OAI.PolycyclicRecognition.pairTransferDefect, OAI.PolycyclicRecognition.pairHeightL2

  5. theorem OAI.PolycyclicRecognition.inverseCoupling_sublinear line 120

    Source documentation

    The actual inverse coupling supplies a sublinear inverse height error with exactly the inverse coefficient. No ergodicity of its finite measure and no second reduced-cohomology hypothesis are required.

    theorem inverseCoupling_sublinear
        {G W : Type*} [Group G] [MetricSpace G] [ProperSpace G]
        [SecondCountableTopology G] [IsTopologicalGroup G] [IsIsometricSMul G G]
        {Λ : ℝ≥0} [MeasurableSpace G] [BorelSpace G]
        [MeasurableSpace (NormalizedPair G Λ)] [BorelSpace (NormalizedPair G Λ)]
        [NormedAddCommGroup W] [NormedSpace ℝ W] [CompleteSpace W] [SecondCountableTopology W]
        [MeasurableSpace W] [BorelSpace W]
        (m : Measure G) [m.IsHaarMeasure] [m.IsMulRightInvariant]
        (μ : Measure (NormalizedPair G Λ)) [IsProbabilityMeasure μ]
        [SMulInvariantMeasure G (NormalizedPair G Λ) μ]
        (A : W ≃L[ℝ] W) (π l : G →* Multiplicative W)
        (hπ : Continuous fun g => (π g).toAdd) (hl : Continuous fun g => (l g).toAdd)
    …

    Mentions: OAI.PolycyclicRecognition.NormalizedPair