Nonpositive variance forces pointwise constancy under a full-support product measure
ProvedNarrowDNF.Fourier.eq_mean_of_variance_le_zerocombinatoricsnarrow-dnf-authors-importprobabilitystoc-2027
Let be finite, , and . Put , and assume
Then every input satisfies
Thus zero variance gives literal pointwise constancy on the entire finite cube. The strict bias assumptions matter: they make every product weight positive, so no input can be excluded by a zero probability mass.
Preamble
import Mathlib.Algebra.BigOperators.GroupWithZero.Finset
import Mathlib.Algebra.BigOperators.Ring.Finset
import Mathlib.Algebra.Order.BigOperators.Group.Finset
import Mathlib.Algebra.Order.BigOperators.Ring.Finset
import Mathlib.Analysis.SpecialFunctions.Sqrt
import Mathlib.Data.Fintype.BigOperators
import Mathlib.Data.Fintype.Pi
import Mathlib.Data.Real.Basic
import Mathlib.Logic.Equiv.Prod
import Mathlib.Tactic.FieldSimp
import Mathlib.Tactic.Linarith
import Mathlib.Tactic.Ring
import Definitions.Def_NarrowDNF_Influence
import Definitions.Def_NarrowDNF_ProductMeasure
namespace NarrowDNF.Fourier
end NarrowDNF.Fourier
namespace NarrowDNF.ProductMeasure
end NarrowDNF.ProductMeasure
noncomputable section
open scoped BigOperators
open NarrowDNF
open NarrowDNF.Fourier
section Cube
variable {ι : Type*} [Fintype ι] [DecidableEq ι]
Formal statement
theorem NarrowDNF.Fourier.eq_mean_of_variance_le_zero {p : ℝ} (hp0 : 0 < p) (hp1 : p < 1)
(f : (ι → Bool) → ℝ) (hf : Influence.variance p f ≤ 0) (x : ι → Bool) :
f x = ProductMeasure.expectation p f := by sorrySource