OAI/Analysis/Brenier/AtomicPaths.lean

GitHub source file

5 declarations, 0 contain sorry, 0 declared as axioms

  1. theorem OAI.Problem358.atomic_potential_path_L2_bound line 18

    Source documentation

    The interior potential estimate for an actual linear motion of the sites. All auxiliary boundedness and joint-selector hypotheses of the time-integration theorem are consequences of the positive masses and the energy bounds.

    theorem atomic_potential_path_L2_bound
        {d : ℕ} {ι : Type*} [Fintype ι] [Nonempty ι]
        [MeasurableSpace ι] [MeasurableSingletonClass ι]
        (K : Set (E d)) [IsFiniteMeasure (uniformMeasure K)]
        (y z : ι → E d) (m : ι → ℝ) (c r : ι → ℝ → ℝ)
        {w κ : ℝ} {R : ℝ≥0}
        (hw : 0 ≤ w) (hκ : 0 ≤ κ) (hm : ∀ i, 0 < m i)
        (hxR : ∀ᵐ x ∂uniformMeasure K, ‖x‖ ≤ R)
        (hc : ∀ i, Measurable (c i)) (hr : ∀ i, Measurable (r i))
        (hdc : ∀ i t, t ∈ Ioo (0 : ℝ) 1 → HasDerivAt (c i) (r i t) t)
        (hinj : ∀ t ∈ Ioo (0 : ℝ) 1,
          Function.Injective (fun i => (1-t) • y i + t • z i))
    …
  2. theorem OAI.Problem358.atomic_potential_path_L2_bound_of_closed_masses line 77

    Source documentation

    The same actual atomic-motion estimate takes the canonical closed-cell mass equations directly. In particular, existence or measurability of prescribed-mass active selectors is not an additional hypothesis on the intercept path.

    theorem atomic_potential_path_L2_bound_of_closed_masses
        {d : ℕ} {ι : Type*} [Fintype ι] [Nonempty ι]
        [MeasurableSpace ι] [MeasurableSingletonClass ι]
        {K : Set (E d)} [IsFiniteMeasure (uniformMeasure K)] (hK : MeasurableSet K)
        (y z : ι → E d) (m : ι → ℝ) (c r : ι → ℝ → ℝ)
        {w κ : ℝ} {R : ℝ≥0}
        (hw : 0 ≤ w) (hκ : 0 ≤ κ) (hm : ∀ i, 0 < m i)
        (hxR : ∀ᵐ x ∂uniformMeasure K, ‖x‖ ≤ R)
        (hc : ∀ i, Measurable (c i)) (hr : ∀ i, Measurable (r i))
        (hdc : ∀ i t, t ∈ Ioo (0 : ℝ) 1 → HasDerivAt (c i) (r i t) t)
        (hinj : ∀ t ∈ Ioo (0 : ℝ) 1,
          Function.Injective (fun i => (1-t) • y i + t • z i))
    …
  3. theorem OAI.Problem358.SupportingMapInterpolation.Supports.lipschitzWith line 124

    Source documentation

    A globally supporting map with bounded values makes its potential Lipschitz. No differentiability or convexity premise is needed for this implication.

    theorem Supports.lipschitzWith {d : ℕ} {u : E d → ℝ} {T : E d → E d}
        (h : Supports u T) {L : ℝ≥0} (hL : ∀ x, ‖T x‖ ≤ L) :
        LipschitzWith L u := by
    …

    Mentions: OAI.Problem358.SupportingMapInterpolation.Supports

  4. theorem OAI.Problem358.one_third_stability_of_finite_potential_estimate line 145

    Source documentation

    The exact compact-target stability conclusion reduces to a uniform potential estimate for finite targets. The supplier only gives supporting maps and their laws: optimality, uniqueness, the one-third interpolation, and passage to arbitrary compactly supported probabilities are all proved here.

    theorem one_third_stability_of_finite_potential_estimate
        {d : ℕ} {K Y : Set (E d)}
        (hK : IsCompact K) (hKconv : Convex ℝ K)
        (hKint : (interior K).Nonempty) (hY : IsCompact Y) (hYne : Y.Nonempty)
        (L : ℝ≥0) (P : ℝ)
        (hpotential : ∀ (μ ν : Measure (E d)),
          IsProbabilityMeasure μ → IsProbabilityMeasure ν →
          (∃ F : Set (E d), F.Finite ∧ F ⊆ Y ∧ IsSupported μ F) →
          (∃ G : Set (E d), G.Finite ∧ G ⊆ Y ∧ IsSupported ν G) →
          ∃ (u v : E d → ℝ) (T S : E d → E d),
            SupportingMapInterpolation.Supports u T ∧
            SupportingMapInterpolation.Supports v S ∧
    …

    Mentions: OAI.Problem358.SupportingMapInterpolation.Supports

  5. theorem OAI.Problem358.one_third_stability_of_atomic_potential_estimate line 216

    Source documentation

    A potential estimate for positive fixed-mass, noncolliding atomic motions implies the full compact-target theorem. The optimal finite coupling, zero-mass deletion, endpoint laws and exact Wasserstein energy are supplied internally.

    theorem one_third_stability_of_atomic_potential_estimate
        {d : ℕ} {K Y : Set (E d)}
        (hK : IsCompact K) (hKconv : Convex ℝ K)
        (hKint : (interior K).Nonempty) (hY : IsCompact Y) (hYne : Y.Nonempty)
        (L : ℝ≥0) (hYL : ∀ z ∈ Y, ‖z‖ ≤ L) (P : ℝ)
        (hpotential : ∀ (n : ℕ), 0 < n →
          ∀ (m : Fin n → ℝ) (y z : Fin n → E d),
          (∀ i, 0 < m i) → (∑ i, m i) = 1 →
          (∀ i, ‖y i‖ ≤ L) → (∀ i, ‖z i‖ ≤ L) →
          (∀ t : ℝ, 0 < t → t < 1 →
            Function.Injective (fun i => (1-t) • y i + t • z i)) →
          ∃ (u v : E d → ℝ) (T S : E d → E d),
    …