bernoulli_powerset_pair_event_prob_eq_product_measure
ProvedPair (decoupled) event-probability → Mathlib product-measure bridge.
The decoupled matrix-completion model evaluates an event on two independent Bernoulli observation sets via the double powerset sum
This theorem states that this bespoke double sum equals the genuine Mathlib product-measure probability
where of on the Bool-indicator space and is the indicator↔Finset bijection. It is the pair analogue of the single-copy bridge bernoulli_powerset_event_prob_eq_product_measure (2d59092a), and it unlocks Mathlib's product-measure / Fubini / independence / conditional-expectation API on the decoupled (two-copy) powerset model — the substrate needed for the de la Peña–Montgomery-Smith order-2 decoupling forward bound on bernoulli_pair_decoupling_spectral_tail_bound_offdiag (9aaf089d).
import Definitions.Def_matrix_completion_bernoulli_measure import Definitions.Def_matrix_completion_neumann import Mathlib.MeasureTheory.Measure.Prod import Mathlib.MeasureTheory.Integral.Prod open MatrixCompletion open scoped BigOperators Classical open MeasureTheory ProbabilityTheory
theorem bernoulli_powerset_pair_event_prob_eq_product_measure
{n1 n2 : ℕ} (p : NNReal) (hp : p ≤ 1)
(Event : Finset (Fin n1 × Fin n2) → Finset (Fin n1 × Fin n2) → Prop) :
bernoulliPairEventProb (p : ℝ) Event
= ((bernMeasure p hp).prod (bernMeasure p hp)).real
{ω | Event (indicatorToFinset ω.1) (indicatorToFinset ω.2)} := by sorry