Documentation

LeanPool.FullyDynamicMatching.FD1D.Bounds

Bounds #

A canonical witness that the fixed-total count state space is nonempty.

Equations
Instances For
    theorem FD1D.HierarchicalDynamics.stationary_expected_hazard_bound {L m : ℕ} (a : ℝ) (ha : 0 < a) (hm : 0 < m) (haL : 200 * ↑L ≤ a) (μ : FiniteLaw (InventoryState (DyadicNode L) m)) (hμ : (kernel a ha hm).IsStationary μ) :
    (μ.expect fun (x : InventoryState (DyadicNode L) m) => stateHazardEnergy a x L) ≤ 206 / (3 * ↑m ^ 2)

    Equation (11) for an actual stationary law of the concrete hierarchical count kernel.

    theorem FD1D.HierarchicalDynamics.existsUnique_stationaryLaw {L m : ℕ} (a : ℝ) (ha : 0 < a) (hm : 0 < m) :

    The concrete count chain has a unique stationary law.

    theorem FD1D.HierarchicalDynamics.finite_average_expected_hazard_bound {L m : ℕ} (a : ℝ) (ha : 0 < a) (hm : 0 < m) (haL : 200 * ↑L ≤ a) (hlogA : 200 * Real.log (1 + ↑m / a) ≤ a) (μ₀ : FiniteLaw (InventoryState (DyadicNode L) m)) (T : ℕ) (hT : 0 < T) :
    (∑ t ∈ Finset.range T, ((kernel a ha hm).iterate t μ₀).expect fun (x : InventoryState (DyadicNode L) m) => stateHazardEnergy a x L) / ↑T ≤ 206 / (3 * ↑m ^ 2) + 1 / ↑T

    Equation (12) for the actual kernel iterates, from an arbitrary initial count law. No invariance assumption is needed for this potential estimate.