Implementation of the lift_eq tactic for kernels. #
This file registers the lifting of kernel expressions for the lift_eq tactic.
Main declarations #
liftKernelExpr: recursive lifting of a kernel expression.liftKernelQ: its typed version, for given measurable equivalences.finisherKernel: the equivalence between an equality of kernels and the equality of their lifts.
def liftKernelQ {x y w : Lean.Level} {X : Q(Type x)} {Y : Q(Type y)} {mX : Q(MeasurableSpace «$X»)} {mY : Q(MeasurableSpace «$Y»)} {X' Y' : Q(Type w)} {mX' : Q(MeasurableSpace «$X'»)} {mY' : Q(MeasurableSpace «$Y'»)} (ex : Q(«$X'» ≃ᵐ «$X»)) (ey : Q(«$Y'» ≃ᵐ «$Y»)) (κ : Lean.Expr) :
Lean.MetaM
((κ : Q(ProbabilityTheory.Kernel «$X» «$Y»)) × (κ' : Q(ProbabilityTheory.Kernel «$X'» «$Y'»)) × Q(«$κ'» = «$κ».lift))
Typed version of the lifting of a kernel κ : Kernel X Y by the registered lifting functions.
Given the measurable equivalences ex : X' ≃ᵐ X and ey : Y' ≃ᵐ Y returns κ, its lift κ' and
the proof of κ' = κ.lift.
Equations
Instances For
Recursive lifting of a kernel e to the universe w. Returns the lift e' together with a
proof of e' = e.lift. Each step is a single application of a lifting lemma, given the lifts
of the subterms.
Equations
- One or more equations did not get rendered due to their size.
Instances For
The finisher for kernels: (κ = η) = (κ' = η') from κ' = κ.lift and η' = η.lift.
Equations
- finisherKernel pl pr = Lean.Meta.mkAppM `ProbabilityTheory.Kernel.lift_congr_of_eq #[pl, pr]