OAI/Geometry/RCD/Sobolev/CutoffErrorEstimates.lean

GitHub source file

4 declarations, 0 contain sorry, 0 declared as axioms

Annular support and heat bounds for the cutoff error

  1. theorem OAI.RegularPointCoordinates.GoodLaplacianCutoff.error_zero_off_annulus line 19

    Source documentation

    The product-rule error lives only in the transition annulus.

    theorem GoodLaplacianCutoff.error_zero_off_annulus {p : X} {r C : ℝ}
        (χ : GoodLaplacianCutoff μ hQ p r C) (g : SobolevSpace μ hQ) :
        ∀ᵐ z ∂μ, z ∉ {z | r ≤ dist p z ∧ dist p z ≤ 2 * r} →
          cutoffSourceDensity μ hQ g χ.sobolev χ.laplacian z = 0 := by
    …

    Mentions: OAI.RegularPointCoordinates.GoodLaplacianCutoff, OAI.RegularPointCoordinates.SobolevSpace, OAI.RegularPointCoordinates.cutoffSourceDensity

  2. theorem OAI.RegularPointCoordinates.abs_integral_mul_le_of_bound_on line 40

    Source documentation

    A kernel bound is needed only where the integrable error is supported.

    theorem abs_integral_mul_le_of_bound_on {f k : X → ℝ} (hf : Integrable f μ)
        {S : Set X} {A : ℝ} (_hA : 0 ≤ A)
        (hzero : ∀ᵐ z ∂μ, z ∉ S → f z = 0) (hk : ∀ z ∈ S, |k z| ≤ A) :
        |∫ z, k z * f z ∂μ| ≤ A * ∫ z, |f z| ∂μ := by
    …
  3. theorem OAI.RegularPointCoordinates.GoodLaplacianCutoff.heat_error_le line 55

    Source documentation

    The cutoff error is controlled by a heat-kernel estimate on its annulus and its finite L1 mass.

    theorem GoodLaplacianCutoff.heat_error_le {p : X} {r C : ℝ}
        (χ : GoodLaplacianCutoff μ hQ p r C) (g : SobolevSpace μ hQ)
        (k : X → ℝ) {A : ℝ} (hA : 0 ≤ A)
        (hk : ∀ z, r ≤ dist p z → dist p z ≤ 2 * r → |k z| ≤ A) :
        |∫ z, k z * cutoffSourceDensity μ hQ g χ.sobolev χ.laplacian z ∂μ| ≤
          A * ∫ z, |cutoffSourceDensity μ hQ g χ.sobolev χ.laplacian z| ∂μ :=
    …

    Mentions: OAI.RegularPointCoordinates.GoodLaplacianCutoff, OAI.RegularPointCoordinates.SobolevSpace, OAI.RegularPointCoordinates.cutoffSourceDensity

  4. theorem OAI.RegularPointCoordinates.LocalHarmonicEnergy.heatGram_le line 71

    Source documentation

    Unit heat mass and the local gradient bound give a uniform upper bound on the actual cutoff Gram, at every positive time.

    theorem LocalHarmonicEnergy.heatGram_le
        (G : LocalHarmonicEnergy μ hQ U M κ p r L v)
        (χ : X →ᵇ ℝ) (hχ : ∀ z, 0 ≤ χ z ∧ χ z ≤ 1)
        (hχs : HasCompactSupport (χ : X → ℝ))
        (hχΩ : tsupport (χ : X → ℝ) ⊆ ball p (2 * r))
        {t : ℝ} (ht : 0 < t) (x : X) :
        ⟪v, cutoffHeatGram μ hQ hk χ.continuous hχs U ht x 0 v⟫_ℝ ≤
          (L : ℝ) ^ 2 * ‖v‖ ^ 2 := by
    …

    Mentions: OAI.RegularPointCoordinates.LocalHarmonicEnergy, OAI.RegularPointCoordinates.cutoffHeatGram