Weak Moment Map for a Bounded Isotropic Law
ProvedFunctionalInequalities.KLS.CEK.exists_momentMap_aux_modularThis is the private auxiliary form of the weak moment-map existence statement.
Let be finite-dimensional with Borel structure, let be an isotropic probability measure, and let . Assume for -almost every . Then there exists such that is convex, -Lipschitz, and normalized by
Furthermore, the tilted Lebesgue measure pushes forward under the gradient to :
Equivalently, ((volume : Measure E).tilted (-ψ)).map (gradient ψ) = μ. The statement claims no additional regularity of .
Superseded upload. This snapshot was replaced by FunctionalInequalities.KLS.CEK.exists_momentMap_aux_readable: open the readable replacement. The replacement preserves the frozen source and mathematical statement, with a compact declaration context. This historical node remains available for existing imports.
import Definitions.Def_KLSModular_FunctionalInequalities_KLS_CEK_MomentMap
noncomputable section
open MeasureTheory Set
open scoped Topology
section
namespace FunctionalInequalities
end FunctionalInequalities
open FunctionalInequalities
section
namespace FunctionalInequalities.KLS
end FunctionalInequalities.KLS
open FunctionalInequalities FunctionalInequalities.KLS
section
namespace FunctionalInequalities.KLS.CEK
end FunctionalInequalities.KLS.CEK
open FunctionalInequalities FunctionalInequalities.KLS FunctionalInequalities.KLS.CEK
open FunctionalInequalities
open FunctionalInequalities.Analysis
open FunctionalInequalities.Probability
open FunctionalInequalities.Probability.MomentMeasure
variable {E : Type*} [NormedAddCommGroup E] [InnerProductSpace ℝ E]
[FiniteDimensional ℝ E] [MeasurableSpace E] [BorelSpace E]
theorem FunctionalInequalities.KLS.CEK.exists_momentMap_aux_modular (μ : ProbabilityMeasure E)
(hμ : KLS.Isotropic μ) (L : NNReal) (hL : ∀ᵐ y ∂(μ : Measure E), ‖y‖ ≤ (L : ℝ)) :
∃ ψ : E → ℝ, ConvexOn ℝ (univ : Set E) ψ ∧ LipschitzWith L ψ ∧
(∫ x, Real.exp (-ψ x) ∂(volume : Measure E)) = 1 ∧
((volume : Measure E).tilted (fun x => -ψ x)).map (gradient ψ) = (μ : Measure E) := by sorry