OAI/NumberTheory/DukePrimeDegree/Entropy/RootPredecessorInformation.lean
2 declarations, 0 contain sorry, 0 declared as axioms
Finite predecessor information on actual root future fibers
- theorem
OAI.DukePrimeDegree.kernel_label_information_eq_of_fiber_equivline 15Source documentation
Recoding on a stationary kernel fiber does not require measurable equality on its measure-valued label.
theorem kernel_label_information_eq_of_fiber_equiv {Y S α β : Type*} [MeasurableSpace Y] [MeasurableSpace S] [MeasurableSpace α] [MeasurableSingletonClass α] [Countable α] [MeasurableSpace β] [MeasurableSingletonClass β] [Countable β] (μ : Measure Y) [IsFiniteMeasure μ] (κ : Kernel Y Y) [IsMarkovKernel κ] (F : Y → S) (hF : Measurable F) (hstat : μ.bind κ=μ) (hself : ∀ᵐ x ∂μ, ∀ᵐ y ∂κ x, κ y=κ x) (hconst : ∀ᵐ x ∂μ, ∀ᵐ y ∂κ x, F y=F x) (P : Y → α) (hP : Measurable P) (B : Y → β) (hB : Measurable B) (hequiv : ∀ x y, F x=F y → (P x=P y ↔ B x=B y)) : conditionalInformation μ P (MeasurableSpace.comap (fun x => (κ x,F x)) inferInstance) =ᵐ[μ] … - theorem
OAI.DukePrimeDegree.EliminationChain.predecessor_root_informationline 53Source documentation
The same finite predecessor symbol used for the ambient entropy bound has exactly the root-information integral on the finer root fibers.
theorem predecessor_root_information (d : ℕ) {P : Finset (Fin (d+1) × Fin (d+1))} {D : Finset (Fin (d+1))} (c : EliminationChain P D) (hP : InitialRootPattern P) (hf : c.finalEntries=allPositiveRoots (d+1)) (μ : Measure (X (d+1))) [IsProbabilityMeasure μ] (hA : ∀ u : LogCoordinates d, μ.map (fun x => rightAction x (diagonalFlow u))=μ) (ell : Fin (d+1) → ℝ) (hell : ∑ k, ell k=0) (Q : X (d+1) → ℕ) (hQ : Measurable Q) (κ : Kernel (X (d+1)) (X (d+1))) [IsMarkovKernel κ] (hstat : μ.bind κ=μ) (hself : ∀ᵐ x ∂μ, ∀ᵐ y ∂κ x, κ y=κ x) (hversion : ∀ᵐ x ∂μ, κ x=c.ambientFutureMeasure hP hf μ ell hell Q 1 x) …Mentions: OAI.DukePrimeDegree.EliminationChain, OAI.DukePrimeDegree.InitialRootPattern, OAI.DukePrimeDegree.allPositiveRoots, OAI.DukePrimeDegree.diagonalFlow