The complete finite product of summable masses has the product total mass
ProvedInfoLossLAR.WorstCase.GaussianPoissonSeries.dependentProductMass_hasSumcountable-productjoint-distributionnormalizationregev
For any source type and any finite number n of coordinates, suppose each coordinate has a nonnegative summable real mass function. Then the complete joint product mass on n-tuples has a HasSum equal to the product of the individual total masses. The mass may depend on the coordinate. This is a statement about the full joint law, with no finite-source or independent-marginal shortcut and no algorithmic claim.
Preamble
import Mathlib set_option autoImplicit false open scoped BigOperators
Formal statement
namespace InfoLossLAR.WorstCase.GaussianPoissonSeries
theorem dependentProductMass_hasSum {Index : Type*} (dimension : ℕ)
(mass : Fin dimension → Index → ℝ)
(hnonnegative : ∀ index value,0 ≤ mass index value) (hsummable : ∀ index,Summable (mass index)) :
HasSum (fun word : Fin dimension → Index => ∏ index,mass index (word index))
(∏ index,∑' value,mass index value) := by sorry
end InfoLossLAR.WorstCase.GaussianPoissonSeriesSource
Exact existing complete proof InfoLossLAR.WorstCase.GaussianPoissonSeries.dependentProductMass_hasSum in Solutions/RegevBridgeDependentProductMass.lean, frozen source SHA 35d70b887158c65e83faa69ff18bac2ef1271fdc33e1b8c9bb95f81b0ada9bb0. Official Lean syntax/info-tree declaration byte ranges are used; only the root binding is renamed to global solution and an unused import closure is removed. No new mathematical argument.