Learning.randomSampling.iIndep_action
Actions are mutually independent.
Learning.randomSampling.iIndep_action.{u_1, u_2, u_3} {๐ : Type u_1} {๐จ : Type u_2} {ฮฉ : Type u_3} {m๐ : MeasurableSpace ๐} {m๐จ : MeasurableSpace ๐จ} {mฮฉ : MeasurableSpace ฮฉ} {ฮผ : MeasureTheory.Measure ๐} [MeasureTheory.IsProbabilityMeasure ฮผ] {P : MeasureTheory.Measure ฮฉ} [MeasureTheory.IsProbabilityMeasure P] {A : โ โ ฮฉ โ ๐} {Y : โ โ ฮฉ โ ๐จ} {env : Environment ๐ ๐จ} (h : IsAlgEnvSeq A Y (randomSampling ฮผ) env P) : ProbabilityTheory.iIndepFun A PLearning.randomSampling.iIndep_action.{u_1, u_2, u_3} {๐ : Type u_1} {๐จ : Type u_2} {ฮฉ : Type u_3} {m๐ : MeasurableSpace ๐} {m๐จ : MeasurableSpace ๐จ} {mฮฉ : MeasurableSpace ฮฉ} {ฮผ : MeasureTheory.Measure ๐} [MeasureTheory.IsProbabilityMeasure ฮผ] {P : MeasureTheory.Measure ฮฉ} [MeasureTheory.IsProbabilityMeasure P] {A : โ โ ฮฉ โ ๐} {Y : โ โ ฮฉ โ ๐จ} {env : Environment ๐ ๐จ} (h : IsAlgEnvSeq A Y (randomSampling ฮผ) env P) : ProbabilityTheory.iIndepFun A P
Code
lemma iIndep_action (h : IsAlgEnvSeq A Y (randomSampling ฮผ) env P) :
iIndepFun A PProof
by
have hA := h.measurable_action
rw [iIndepFun_nat_iff_forall_indepFun (by fun_prop)]
intro n
have map_eq := (h.hasCondDistrib_action n).map_eq
simp only [randomSampling_policy, Measure.compProd_const] at map_eq
have law_eq : P.map (A (n + 1)) = ฮผ := (hasLaw_action h (n + 1)).map_eq
rw [โ law_eq, โ indepFun_iff_map_prod_eq_prod_map_map] at map_eq
ยท change A (n + 1) โแตข[P] (fun (f : Iic n โ ๐ ร ๐จ) โฆ (fun i โฆ (f i).1))โ (history A Y n)
refine map_eq.symm.comp measurable_id (by fun_prop)
ยท exact (h.measurable_history n).aemeasurable
ยท exact (h.measurable_action (n + 1)).aemeasurableActions: Source ยท Open Issue
Meaning last changed in v4.34.0-rc2-1-g439785b (2026-08-23), the 4th recorded change.
Self-contained, with its dependencies inlined and proofs replaced by sorry: download the raw file ยท open it in the Lean web editor.
Dependency graph
Audit surface: 5 project declarations, 43 external constants
โ Proved: no sorry anywhere in its closure
This is the tool's own reading of one build's recorded axioms, and it is not robust against an author who wants it to pass. Checking meant to be relied on should go through Comparator, which replays the proof through the kernel from an export against an explicit list of permitted axioms.