OAI/GroupTheory/PolycyclicRecognition/Analysis/StationarySlope.lean

GitHub source file

15 declarations, 0 contain sorry, 0 declared as axioms

Averaging the stationary height cocycle

The normalized pair space carries a measure-preserving action. Its height evaluation is a cocycle for inverse pullback. We construct that action on the actual space L², and show that integration gives an additive slope. No reduced-cohomology assertion is assumed in this construction.

  1. def OAI.PolycyclicRecognition.stationaryPullback line 32

    Source documentation

    Inverse pullback is the left representation associated to a left action.

    def stationaryPullback (g : G) : Lp W 2 μ →+ Lp W 2 μ :=
    …
  2. theorem OAI.PolycyclicRecognition.stationaryPullback_one line 35
  3. theorem OAI.PolycyclicRecognition.stationaryPullback_mul line 41
  4. def OAI.PolycyclicRecognition.stationaryL2Action line 50

    Source documentation

    The isometric additive action on the genuine L² space.

    def stationaryL2Action : DistribMulAction G (Lp W 2 μ) where
    …
  5. theorem OAI.PolycyclicRecognition.stationaryPullback_const_finite line 61

    Source documentation

    Constants are invariant for a finite invariant measure, without a probability normalization.

    theorem stationaryPullback_const_finite [IsFiniteMeasure μ] (g : G) (w : W) :
        stationaryPullback μ g (Lp.const 2 μ w) = Lp.const 2 μ w := by
    …

    Mentions: OAI.PolycyclicRecognition.stationaryPullback

  6. instance OAI.PolycyclicRecognition.stationaryL2Isometric line 69
  7. theorem OAI.PolycyclicRecognition.continuous_stationaryPullback line 82

    Source documentation

    Strong continuity follows from the actual continuous action on the underlying measure space and density of simple functions in L².

    theorem continuous_stationaryPullback :
        Continuous (fun z : G × Lp W 2 μ => stationaryPullback μ z.1 z.2) := by
    …

    Mentions: OAI.PolycyclicRecognition.stationaryPullback

  8. theorem OAI.PolycyclicRecognition.isCocycle_toL2 line 96

    Source documentation

    Pointwise measurable height cocycles give L² cocycles without selecting pointwise representatives of the resulting equivalence classes.

    theorem isCocycle_toL2 (b : G → X → W) (hb : ∀ g, MemLp (b g) 2 μ)
        (hc : ∀ g h x, b (g * h) x = b g x + b h (g⁻¹ • x)) :
        IsCocycle (fun g => (hb g).toLp (b g)) := by
    …

    Mentions: OAI.PolycyclicRecognition.IsCocycle

  9. instance OAI.PolycyclicRecognition.stationaryL2SMulComm line 113
  10. theorem OAI.PolycyclicRecognition.norm_integral_le_l2 line 119

    Source documentation

    The averaging operator is bounded by one on a probability space.

    theorem norm_integral_le_l2 [IsProbabilityMeasure μ] (v : Lp W 2 μ) :
        ‖∫ x, v x ∂μ‖ ≤ ‖v‖ := by
    …
  11. def OAI.PolycyclicRecognition.stationaryL2Mean line 132

    Source documentation

    Averaging as an actual bounded linear operator on L².

    def stationaryL2Mean [IsProbabilityMeasure μ] : Lp W 2 μ →L[ℝ] W :=
    …
  12. theorem OAI.PolycyclicRecognition.stationaryL2Mean_pullback line 148

    Source documentation

    The mean is invariant under the pullback representation.

    theorem stationaryL2Mean_pullback [IsProbabilityMeasure μ] (g : G) (v : Lp W 2 μ) :
        stationaryL2Mean μ (stationaryPullback μ g v) = stationaryL2Mean μ v := by
    …

    Mentions: OAI.PolycyclicRecognition.stationaryL2Mean, OAI.PolycyclicRecognition.stationaryPullback

  13. def OAI.PolycyclicRecognition.stationaryMean line 159

    Source documentation

    The average is a genuine homomorphism; invariance removes the translated second summand in the cocycle equation.

    def stationaryMean (b : G → X → W) (hb : ∀ g, Integrable (b g) μ)
        (hc : ∀ g h x, b (g * h) x = b g x + b h (g⁻¹ • x)) :
        G →* Multiplicative W where
    …
  14. theorem OAI.PolycyclicRecognition.integral_centered_stationaryMean line 180

    Source documentation

    On a probability space the centered pointwise cocycle has zero mean.

    theorem integral_centered_stationaryMean [IsProbabilityMeasure μ]
        (b : G → X → W) (hb : ∀ g, Integrable (b g) μ)
        (hc : ∀ g h x, b (g * h) x = b g x + b h (g⁻¹ • x)) (g : G) :
        (∫ x, b g x - (stationaryMean μ b hb hc g).toAdd ∂μ) = 0 := by
    …

    Mentions: OAI.PolycyclicRecognition.stationaryMean

  15. theorem OAI.PolycyclicRecognition.stationary_fixed_eq_const line 189

    Source documentation

    Ergodicity of the action, rather than an assumption about its Hilbert representation, identifies all invariant L² vectors with constants.

    theorem stationary_fixed_eq_const
        {G X W : Type*} [Group G] [MeasurableSpace X] [MulAction G X]
        [MeasurableConstSMul G X] [NormedAddCommGroup W] [SecondCountableTopology W]
        (μ : Measure X) [IsProbabilityMeasure μ] [ErgodicSMul G X μ] (v : Lp W 2 μ)
        (hfix : ∀ g : G, stationaryPullback μ g v = v) :
        ∃ c : W, v = Lp.const 2 μ c := by
    …

    Mentions: OAI.PolycyclicRecognition.stationaryPullback