Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

Quadrature bound for the variation majorant over E2

Open
LL_DensitySlope517.leaf_two_variation_quadrature

by sthubbar · Nov 17, 2026 · Mathlib 0df444a (Lean v4.33.1)

number-theory

Let ε\varepsilonε be real with 0<ε≤10−80<\varepsilon\le10^{-8}0<ε≤10−8. Let F(a,b)=1/(ab(99/100−a−b)(1−a−b))F(a,b)=1/\bigl(ab(99/100-a-b)(1-a-b)\bigr)F(a,b)=1/(ab(99/100−a−b)(1−a−b)) if a+3b≥1a+3b\ge1a+3b≥1, F(a,b)=(107/200)/(ab3)F(a,b)=(107/200)/(ab^3)F(a,b)=(107/200)/(ab3) if 99/100≤a+3b<199/100\le a+3b<199/100≤a+3b<1, and F(a,b)=(7/200)/(ab3)F(a,b)=(7/200)/(ab^3)F(a,b)=(7/200)/(ab3) if a+3b<99/100a+3b<99/100a+3b<99/100. Then the iterated interval integral ∫1/101/2∫1/101/2G(a,b) db da\int_{1/10}^{1/2}\int_{1/10}^{1/2}G(a,b)\,db\,da∫1/101/2​∫1/101/2​G(a,b)dbda is at most 222, where G(a,b)=F(a,b)G(a,b)=F(a,b)G(a,b)=F(a,b) when (a,b)(a,b)(a,b) lies in the set E2E_2E2​ given by the predicate above with parameter ε\varepsilonε, and G(a,b)=0G(a,b)=0G(a,b)=0 otherwise.

Preamble
import Mathlib
import Definitions.Def_LL_DensitySlope517_lowerTwoDiscarded

open Classical
Formal statement
namespace LL_DensitySlope517
theorem leaf_two_variation_quadrature {ε : ℝ} (hε : 0 < ε) (hε8 : ε ≤ 1/10^8) :
    (∫ a in (1/10:ℝ)..(1/2:ℝ), ∫ b in (1/10:ℝ)..(1/2:ℝ),
      (if lowerTwoDiscarded ε a b then
        (if 1 ≤ a+3*b then 1/(a*b*(99/100-a-b)*(1-a-b))
          else if 99/100 ≤ a+3*b then (107/200)/(a*b^3) else (7/200)/(a*b^3))
      else 0)) ≤ 2 := by sorry
end LL_DensitySlope517
Source
Baker, Harman and Pintz 1997, Section 5, page 51, display 5.17; the explicit majorant of the variation of the E2E_2E2​ integrand used to bound the slope of the lower density near s=1s=1s=1.
Read-back

What the Lean code literally says, in plain math · claude-haiku

Declaration: theorem leaf_two_variation_quadrature in namespace LL_DensitySlope517. It is a theorem, and its statement is a single inequality about a real-valued integral.

Binders and hypotheses. The theorem takes a real number ε\varepsilonε (implicit binder {ε : ℝ}) with two hypotheses: ε>0\varepsilon > 0ε>0 (hε) and ε≤1/108\varepsilon \le 1/10^8ε≤1/108 (hε8). The file also opens Classical, so the if ... then ... else below uses classical decidability. The file imports a definition lowerTwoDiscarded from another module, Definitions.Def_LL_DensitySlope517_lowerTwoDiscarded. That definition is not expanded here, so the readback cannot state what it says. It is a predicate lowerTwoDiscarded(ε,a,b)\texttt{lowerTwoDiscarded}(\varepsilon, a, b)lowerTwoDiscarded(ε,a,b) that the integrand tests at each point (a,b)(a,b)(a,b). The truth of the statement depends on which (a,b)(a,b)(a,b) satisfy it.

Integrand. Define, for real a,ba, ba,b:

F(a,b)={G(a,b)if lowerTwoDiscarded(ε,a,b) holds,0otherwise,F(a,b) = \begin{cases} G(a,b) & \text{if } \texttt{lowerTwoDiscarded}(\varepsilon,a,b) \text{ holds},\\ 0 & \text{otherwise}, \end{cases}F(a,b)={G(a,b)0​if lowerTwoDiscarded(ε,a,b) holds,otherwise,​

where G(a,b)G(a,b)G(a,b) is given by three cases:

G(a,b)={1a b (99100−a−b)(1−a−b)if a+3b≥1,107/200a b3if a+3b<1 and a+3b≥99100,7/200a b3if a+3b<99100.G(a,b) = \begin{cases} \dfrac{1}{a\,b\,\left(\tfrac{99}{100}-a-b\right)\left(1-a-b\right)} & \text{if } a+3b \ge 1,\\[2ex] \dfrac{107/200}{a\,b^{3}} & \text{if } a+3b < 1 \text{ and } a+3b \ge \tfrac{99}{100},\\[2ex] \dfrac{7/200}{a\,b^{3}} & \text{if } a+3b < \tfrac{99}{100}. \end{cases}G(a,b)=⎩⎨⎧​ab(10099​−a−b)(1−a−b)1​ab3107/200​ab37/200​​if a+3b≥1,if a+3b<1 and a+3b≥10099​,if a+3b<10099​.​

Conclusion. The statement is

∫1/101/2(∫1/101/2F(a,b) db)da  ≤  2.\int_{1/10}^{1/2} \left( \int_{1/10}^{1/2} F(a,b)\, db \right) da \;\le\; 2.∫1/101/2​(∫1/101/2​F(a,b)db)da≤2.

The inner integral runs over bbb and the outer over aaa, both on [1/10,1/2][1/10, 1/2][1/10,1/2] as Lean interval integrals. The conclusion is an upper bound of 2, not an equality. Nothing else is asserted. In particular, the theorem does not say the integral is finite or positive.

Concerns for the auditor.

  1. Unexpanded predicate. The region of integration is determined by lowerTwoDiscarded, which is imported and not shown in this file. Its exact meaning must be checked in its own definition file before the statement can be judged faithful. Any reading of this statement is incomplete until that is done.

  2. Lean junk-value convention. In Lean, an integral of a function that is not integrable on the interval is defined to be 000. So if FFF is not Lebesgue integrable on the square (see item 3), the conclusion holds trivially as 0≤20 \le 20≤2. A reader expecting a substantive bound should check integrability separately. The statement itself does not assert it.

  3. Apparent singularities inside the domain. The first branch has denominator (99100−a−b)(1−a−b)\left(\tfrac{99}{100}-a-b\right)\left(1-a-b\right)(10099​−a−b)(1−a−b). The first branch is active only when a+3b≥1a+3b \ge 1a+3b≥1. For a,b∈[1/10,1/2]a, b \in [1/10, 1/2]a,b∈[1/10,1/2], the line a+b=99/100a+b = 99/100a+b=99/100 meets the region a+3b≥1a+3b \ge 1a+3b≥1 (for example, b∈[0.49,0.5]b \in [0.49, 0.5]b∈[0.49,0.5] with a=0.99−ba = 0.99 - ba=0.99−b). Along that segment the integrand has a simple pole, and a simple pole along a line is not Lebesgue integrable in two dimensions. So unless lowerTwoDiscarded excludes that segment, the integrand may fail to be integrable, and then the conclusion is vacuous by item 2. The line a+b=1a+b=1a+b=1 meets the domain only at the corner (1/2,1/2)(1/2, 1/2)(1/2,1/2), which has measure zero.

  4. Other points. The denominators aaa, bbb, and b3b^3b3 are positive on the domain, since a,b≥1/10a, b \ge 1/10a,b≥1/10, so those factors cause no trouble. The branch boundaries are consistent: the three cases cover all real a+3ba+3ba+3b, with the first case taking a+3b≥1a+3b \ge 1a+3b≥1, the second 99/100≤a+3b<199/100 \le a+3b < 199/100≤a+3b<1, and the third a+3b<99/100a+3b < 99/100a+3b<99/100.

  5. Hypotheses. ε\varepsilonε enters only through lowerTwoDiscarded. The hypotheses 0<ε≤10−80 < \varepsilon \le 10^{-8}0<ε≤10−8 therefore constrain the region only indirectly. If lowerTwoDiscarded is not monotone or otherwise sensitive to ε\varepsilonε in the way the hypotheses suggest, the bound may not reflect the intended dependence on ε\varepsilonε. This requires the imported definition.

  6. Strength. The bound is one-sided and uses the constant 2 with no ε\varepsilonε dependence in the conclusion. A reader expecting a bound that depends on ε\varepsilonε should note that the stated bound is uniform in ε\varepsilonε.

Summary. The theorem says that a particular double interval integral over [1/10,1/2]2[1/10,1/2]^2[1/10,1/2]2, restricted to the set where the imported predicate lowerTwoDiscarded holds, is at most 2. The integrand has an apparent non-integrable singularity along a+b=99/100a+b = 99/100a+b=99/100 inside the domain. The result may therefore be vacuously true (integral 0 by Lean convention) if that singularity is not excluded by the predicate. Faithfulness cannot be judged without expanding lowerTwoDiscarded.

readback_model: claude-haiku-5-5

Human review
  • Endorsed by Shuze Chen · Nov 17, 2026

    Confirmed by the moderator at approval.

  • Endorsed by sthubbar · Nov 17, 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