Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

§4.3, triangle-inequality failure

Open
QAlgorithms.Papers.Independent.Intakenext100.P45.triangleFailure

by Gohan · Oct 31, 2026 · Mathlib 0df444a (Lean v4.33.1)

prove2quantum

For Bernoulli laws pε=(ε,1−ε)p_\varepsilon=(\varepsilon,1-\varepsilon)pε​=(ε,1−ε), qε=(1−ε,ε)q_\varepsilon=(1-\varepsilon,\varepsilon)qε​=(1−ε,ε) and u=(1/2,1/2)u=(1/2,1/2)u=(1/2,1/2), there is δ∈(0,1/2]\delta\in(0,1/2]δ∈(0,1/2] such that every 0<ε<δ0<\varepsilon<\delta0<ε<δ gives finite Chernoff distances and

ξ(pε,qε)>ξ(pε,u)+ξ(u,qε).\sqrt{\xi(p_\varepsilon,q_\varepsilon)}>\sqrt{\xi(p_\varepsilon,u)}+\sqrt{\xi(u,q_\varepsilon)}.ξ(pε​,qε​)​>ξ(pε​,u)​+ξ(u,qε​)​.

Thus even the square root of the Chernoff distance need not satisfy the triangle inequality. Positivity of ε\varepsilonε keeps every term finite.

Preamble
import Definitions.Def_QAlgorithms_Papers_Independent_Intakenext100_P45
Formal statement
noncomputable section

namespace QAlgorithms.Papers.Independent.Intakenext100.P45

open QAlgorithms Filter

open scoped Matrix ComplexConjugate ComplexOrder ENNReal BigOperators Topology

open QAlgorithms.Papers.Independent.Information100.P07
  (support positivePart positiveProjection IsProjection)

open QAlgorithms.Papers.Independent.Information100.P56 (IsLaw)

open QAlgorithms.Papers.Independent.Information100.P60 (Mat copies)

open QAlgorithms.Papers.Independent.Dependency50.P28 (logNonnegativeExtended relativeEntropy)

open QAlgorithms.Papers.Independent.Claimed50.P15 (positiveMatrixRpow)

attribute [local instance] Classical.propDecidable Classical.decEq

variable {I J : Type} [Fintype I] [DecidableEq I] [Fintype J] [DecidableEq J]

/-- arXiv:0708.4282v1, PDF p.17: §4.3, triangle-inequality failure. -/
theorem triangleFailure  : ∃ delta : ℝ, 0 < delta ∧ delta ≤ 1/2 ∧
    ∀ eps : ℝ, 0 < eps → eps < delta →
    let p : Fin 2 → ℝ := ![eps,1-eps]
    let q : Fin 2 → ℝ := ![1-eps,eps]
    let u : Fin 2 → ℝ := ![1/2,1/2]
    classicalDistance p q ≠ ⊤ ∧ classicalDistance p u ≠ ⊤ ∧
    classicalDistance u q ≠ ⊤ ∧
    Real.sqrt (classicalDistance p u).toReal + Real.sqrt (classicalDistance u q).toReal <
      Real.sqrt (classicalDistance p q).toReal := by sorry

end QAlgorithms.Papers.Independent.Intakenext100.P45

end
Source
Asymptotic Error Rates in Quantum Hypothesis Testing, arXiv:0708.4282v1, PDF p. 17, §4.3, triangle-inequality failure
Read-back

What the Lean code literally says, in plain math · gpt-6.1-sol

There exists a real δ\deltaδ with 0<δ≤120<\delta\leq\frac120<δ≤21​ such that, for every real ε\varepsilonε satisfying 0<ε<δ0<\varepsilon<\delta0<ε<δ, the two-coordinate real functions

p=(ε,1−ε),q=(1−ε,ε),u=(12,12)p=(\varepsilon,1-\varepsilon),\qquad q=(1-\varepsilon,\varepsilon),\qquad u=\left(\tfrac12,\tfrac12\right)p=(ε,1−ε),q=(1−ε,ε),u=(21​,21​)

have the following properties. For real functions v,wv,wv,w on these two coordinates, define

c(v,w)=inf⁡0≤s≤1∑i=01vi 1−swis,C(v,w)={−log⁡c(v,w),c(v,w)>0,+∞,c(v,w)≤0.c(v,w)=\inf_{0\leq s\leq1}\sum_{i=0}^{1}v_i^{\,1-s}w_i^s, \qquad C(v,w)= \begin{cases} -\log c(v,w),&c(v,w)>0,\\ +\infty,&c(v,w)\leq0. \end{cases}c(v,w)=0≤s≤1inf​i=0∑1​vi1−s​wis​,C(v,w)={−logc(v,w),+∞,​c(v,w)>0,c(v,w)≤0.​

The powers are real powers, with 00=10^0=100=1, and the logarithm is natural. Then C(p,q)C(p,q)C(p,q), C(p,u)C(p,u)C(p,u), and C(u,q)C(u,q)C(u,q) are each unequal to +∞+\infty+∞, and

toReal⁡(C(p,u))+toReal⁡(C(u,q))<toReal⁡(C(p,q)).\sqrt{\operatorname{toReal}(C(p,u))} +\sqrt{\operatorname{toReal}(C(u,q))} < \sqrt{\operatorname{toReal}(C(p,q))}.toReal(C(p,u))​+toReal(C(u,q))​<toReal(C(p,q))​.

Here toReal⁡\operatorname{toReal}toReal returns the underlying real value for a finite extended real and returns zero at either infinity, and the real square root is zero on negative inputs.

Human review
  • Endorsed by Shuze Chen · Oct 31, 2026

    Confirmed by the moderator at approval.

  • Endorsed by Gohan · Oct 31, 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