Second moment of a nonnegative Lipschitz endpoint cost
ProvedDerivativeChainContinuation.uniform_endpoint_second_momentcontinuation-20261026derivative-chainsreal-transport
Let , with , and let satisfy . If and the mean cost is at most times the root-mean-square position, then
This combines a variance estimate with the mean transport bound to obtain the constant .
Preamble
import Mathlib import Definitions.Def_DerivativeChainPrivate_foundations open DerivativeChainPrivate set_option autoImplicit false set_option backward.isDefEq.respectTransparency false open Polynomial open scoped BigOperators
Formal statement
theorem DerivativeChainContinuation.uniform_endpoint_second_moment (n : ℕ) (hn : 0 < n)
(x f : Fin n → ℝ) (hcenter : ∑ i, x i = 0)
(hcost : ∀ i, 0 ≤ f i)
(hlip : ∀ i j, |f i-f j| ≤ |x i-x j|)
(B : ℝ) (hB : 0 ≤ B)
(hmean : (∑ i, f i)/(n : ℝ) ≤ B*Real.sqrt ((∑ i, (x i)^2)/(n : ℝ))) :
(∑ i, (f i)^2)/(n : ℝ) ≤ (1+B^2)*((∑ i, (x i)^2)/(n : ℝ)) := by sorrySource