The raw configuration inequality with regularisation error
ProvedApery.energy_rawLet , , , let , and let be pairwise distinct real numbers. With the potential of the comparison measure, the total regularisation error, and the logarithmic energy of Appendix A,
This is the raw configuration inequality of the source (Apery/EnergyRaw.lean), the starting point of the proof of the logarithmic-energy inequality (6.9): it is obtained by inserting the regularised pair potential on a circle, using Lemmas 6.1 and 6.2 (through the Arcsine/Gaussian kernel decomposition) and letting the circle radius tend to infinity. The dependence on the configuration is only through the one-point terms .
Formalization Note The statement uses the platform definitions Definitions.Def_Zeta5_SourceEnergy (for , ), Definitions.Def_Zeta5_SourceConstants (for ) and Definitions.Def_Zeta5_SourceMeasure (for ); the configuration is indexed by Fin (37 * n).
import Mathlib import Definitions.Def_Zeta5_SourceConstruction import Definitions.Def_Zeta5_SourceValue import Definitions.Def_Zeta5_SourceConstants import Definitions.Def_Zeta5_SourceMeasure import Definitions.Def_Zeta5_SourceAnalytic import Definitions.Def_Zeta5_SourceEnergy open Polynomial Filter Topology Finset MeasureTheory
namespace Apery
theorem energy_raw (n : ℕ) (hn : 0 < n) (t : Fin (37 * n) → ℝ) (hinj : Function.Injective t)
{ε : ℝ} (hε : 0 < ε) :
∑ i, ∑ i', Real.log |t i - t i'| ≤
-(37 * n : ℝ) * Real.log ε + 2 * Kr n * ∑ i, (Uρ (t i) + Err ε (t i)) - Kr n ^ 2 * Irho := by sorry
end Apery