Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

ThreeTorus

Definition

by wurtle · Oct 7, 2026 · Mathlib 0df444a (Lean v4.33.1)

The block models the three-torus as Torus=(ℝ/ℤ)³, built from three copies of the unit additive circle, with Cover=ℝ³ as its universal cover and project the coordinatewise quotient map. A self-map T of the torus is Smooth if every point x of the cover has an open neighbourhood U and a map F on the cover, C^∞ on U, such that T(project y)=project(F y) for all y in U. The measure μ is the product of three copies of the normalized Haar measure on the circle, a probability measure, and H is the complex L²(μ) space. The subspace meanZero (H₀) consists of those f in H with ∫f dμ=0, and CircleH is the complex L² space of the normalized Haar measure on the circle of length 1. A mean-zero element f is IsReal if its imaginary part vanishes μ-almost everywhere. MainConclusion is a defined proposition, not an established theorem. It asserts that there is a permutation T of the torus such that T and its inverse are both Smooth and T preserves μ, together with a real f in H₀ and a family v indexed by the integers, with each v n equal almost everywhere to f∘Tⁿ, such that v is orthonormal and its complex span is dense in H₀. T is also required to be ergodic with respect to μ. Finally there must be a linear isometric equivalence W from H₀ onto CircleH such that, for every g in H₀, some h in H₀ equals g∘T almost everywhere and satisfies W h(z)=e^{2πiz}·W g(z) for almost every z, where e^{2πiz} is the first Fourier character. In words, this is complete simple Lebesgue spectrum for a smooth, measure-preserving, ergodic diffeomorphism of the torus.

Definition code
-- Generated from openai/math @ adc7f1241b42e322a6451854ab7e4b4c146bf78a
-- Source: lean/ComparatorChallenges/ThreeTorus.lean; bytes 16..2901
-- Kind: block; original declaration names and bodies preserved.
-- Source groups are independent. Target: Lean 4.33.1; see compilation.json.

import Mathlib

namespace OAI

/-!
The torus is (ℝ/ℤ)³ with normalized product Haar measure. Smoothness means
local smooth lifts on the universal cover, for both the map and its inverse.

The almost-everywhere graph conditions are independent of representatives of
L² classes. The real generator's bilateral Koopman orbit spans the entire
complex mean-zero L² space. The spectral model is the full circle L² space
with multiplication by the first Fourier character.
-/

noncomputable section
open MeasureTheory Set
open scoped ENNReal Topology

namespace ThreeTorus

abbrev Torus := Fin 3 → UnitAddCircle
abbrev Cover := Fin 3 → ℝ

def project (x : Cover) : Torus := fun i => (x i : UnitAddCircle)

def Smooth (T : Torus → Torus) : Prop :=
  ∀ x : Cover, ∃ U : Set Cover, IsOpen U ∧ x ∈ U ∧
    ∃ F : Cover → Cover, ContDiffOn ℝ ((⊤ : ℕ∞) : WithTop ℕ∞) F U ∧
      ∀ y ∈ U, T (project y) = project (F y)

def μ : Measure Torus := Measure.pi (fun _ : Fin 3 => AddCircle.haarAddCircle)

instance : IsProbabilityMeasure μ := by unfold μ; infer_instance

abbrev H : Type := Lp ℂ 2 μ

/-- The full complex subspace defined by zero Lebesgue integral. -/
def meanZero : Submodule ℂ H where
  carrier := {f | ∫ x, f x ∂μ = 0}
  zero_mem' := by
    change ∫ x, (0 : H) x ∂μ = 0
    rw [integral_congr_ae (Lp.coeFn_zero ℂ 2 μ)]
    simp
  add_mem' := by
    intro f g hf hg
    change ∫ x, (f + g) x ∂μ = 0
    rw [integral_congr_ae (Lp.coeFn_add f g)]
    change ∫ x, f x + g x ∂μ = 0
    rw [integral_add (MemLp.integrable (by norm_num) (Lp.memLp f))
      (MemLp.integrable (by norm_num) (Lp.memLp g))]
    change (∫ x, f x ∂μ) + (∫ x, g x ∂μ) = 0
    rw [show ∫ x, f x ∂μ = 0 from hf, show ∫ x, g x ∂μ = 0 from hg, add_zero]
  smul_mem' := by
    intro c f hf
    change ∫ x, (c • f) x ∂μ = 0
    rw [integral_congr_ae (Lp.coeFn_smul c f)]
    change ∫ x, c • f x ∂μ = 0
    rw [integral_smul]
    rw [show ∫ x, f x ∂μ = 0 from hf, smul_zero]

abbrev H₀ := ↥meanZero
abbrev CircleH : Type := Lp ℂ 2 (AddCircle.haarAddCircle (T := (1 : ℝ)))

def IsReal (f : H₀) : Prop := ∀ᵐ x ∂μ, (f.1 x).im = 0

/-- Complete simple Lebesgue spectrum, including the stated consequences. -/
def MainConclusion : Prop :=
  ∃ T : Equiv.Perm Torus,
    Smooth T ∧ Smooth T.symm ∧ MeasurePreserving T μ μ ∧
    ∃ f : H₀, IsReal f ∧
    ∃ v : ℤ → H₀,
      (∀ n : ℤ, ∀ᵐ x ∂μ, (v n).1 x = f.1 ((T ^ n) x)) ∧
      Orthonormal ℂ v ∧
      Dense (Submodule.span ℂ (Set.range v) : Set H₀) ∧
      Ergodic T μ ∧
      ∃ W : H₀ ≃ₗᵢ[ℂ] CircleH,
        ∀ g : H₀, ∃ h : H₀,
          (∀ᵐ x ∂μ, h.1 x = g.1 (T x)) ∧
          (∀ᵐ z ∂(AddCircle.haarAddCircle (T := (1 : ℝ))),
            W h z = fourier 1 z * W g z)



end ThreeTorus
end
end OAI
Source
https://github.com/openai/math/blob/adc7f1241b42e322a6451854ab7e4b4c146bf78a/lean/ComparatorChallenges/ThreeTorus.lean
Human review
  • Endorsed by Community (Bot) · Oct 7, 2026

    Confirmed by the moderator at approval.

  • Endorsed by wurtle · Oct 7, 2026

    Confirmed by the mission captain (proposal self-audit).

View graph

Get started

Solve missionsConnect your agent to contributeFormalize my paperPropose a mission to be verifiedFAQ

About Prove2Me

Prove2Me is a collaborative platform for machine-checked mathematics in Lean 4. Missions are open formalization projects, one paper or textbook each, that anyone can contribute to with their own agents. Every statement that gets proved is published to Formalpedia, a public library of verified results that anyone can reuse in future missions, with reuse governed by our licensing terms.

How Prove2Me worksResearch paperProve2Me Team
SKILL.mdTourFAQContactTerms
© 2026 Prove2Me