Equation (9): conditional maximum-threshold comparison
ProvedMinBuying.affiliated_threshold_comparisonFor an exchangeable affiliated nonnegative joint probability law, a coordinate , and thresholds with positive probability that the maximum is at least , the conditional probability of given that the maximum is at least is at least its conditional probability given that the maximum is at least . Affiliation is the actual density-or-PMF condition.
import Definitions.Def_MinBuying_Symmetry
namespace MinBuying
/-- Equation (9), obtained in the source through affiliation of order
statistics. Conditioning events have positive probability; this is a theorem
obligation rather than an additional definition of affiliation. -/
theorem affiliated_threshold_comparison (n : ℕ) (hn : 0 < n) (μ : JointLaw n)
(hμ : ValidLaw μ) (hex : Exchangeable μ) (haff : Affiliated μ)
(i : Fin n) (a b : ℝ) (hab : a ≤ b)
(hb : 0 < μ (upperMaximumEvent b)) :
(μ (coordinateBelowAndUpperMaximumEvent i a b)).toReal /
(μ (upperMaximumEvent b)).toReal ≤
(μ (coordinateBelowAndUpperMaximumEvent i a a)).toReal /
(μ (upperMaximumEvent a)).toReal := by
sorry
end MinBuying
Read-back
What the Lean code literally says, in plain math · gpt-6-astra
For every integer , let be a probability measure on such that all coordinates are nonnegative simultaneously -almost surely. Assume is invariant under every map for a permutation of the coordinates, and assume one of the following alternatives holds: either has an everywhere nonnegative, Borel-measurable real density with respect to Lebesgue measure on , satisfying for every , or is the measure of a probability mass function on whose real point masses satisfy for every . Here and mean coordinatewise maximum and minimum, and the mass-function alternative does not require finite support. For every index and every pair of real numbers with , put and . If , then , with all event masses read as finite real numbers. The coordinate-below test is strict and the maximum-threshold test is weak. The thresholds need not be nonnegative, and equality is allowed. Since and is a probability measure, the given positive-mass hypothesis makes both displayed denominators finite and strictly positive.
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.