Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

§5.2, conditional rate shifts (43)

Open
QAlgorithms.Papers.Independent.Intakenext100.P45.conditionalRateShifts

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

prove2quantum

For nonorthogonal finite laws and a null-admissible nonrandomized test sequence, original error-rate limits exist exactly when the corresponding conditional limits exist. Their values satisfy

αR=−log⁡ψ0+α~R,βR=−log⁡ψ1+β~R.\alpha_R=-\log\psi_0+\widetilde\alpha_R,\qquad\beta_R=-\log\psi_1+\widetilde\beta_R.αR​=−logψ0​+αR​,βR​=−logψ1​+β​R​.

Both ψ0,ψ1\psi_0,\psi_1ψ0​,ψ1​ are positive. The shifts are finite and apply also to infinite rate limits.

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.23: §5.2, conditional rate shifts (43). -/
theorem conditionalRateShifts (p q : I → ℝ) (hp : IsLaw p) (hq : IsLaw q)
    (h0 : 0 < overlapMass p q) (h1 : 0 < overlapMass q p)
    (f : (k : ℕ) → ObservableWords p q (k+1) → Bool)
    (hf : ∀ k, NullAdmissible p q (f k)) :
    (∀ a : EReal,
      HasRate (fun k => restrictedAlpha p q (f k)) a ↔
      HasRate (fun k => observableAlpha p q (f k))
        (((-Real.log (overlapMass p q) : ℝ) : EReal)+a)) ∧
    (∀ b : EReal,
      HasRate (fun k => restrictedBeta p q (f k)) b ↔
      HasRate (fun k => observableBeta p q (f k))
        (((-Real.log (overlapMass q p) : ℝ) : EReal)+b)) ∧
    ((∃ a, HasRate (fun k => observableAlpha p q (f k)) a) ↔
     (∃ a, HasRate (fun k => restrictedAlpha p q (f k)) a)) ∧
    ((∃ b, HasRate (fun k => observableBeta p q (f k)) b) ↔
     (∃ b, HasRate (fun k => restrictedBeta p q (f k)) b)) := by sorry

end QAlgorithms.Papers.Independent.Intakenext100.P45

end
Source
Asymptotic Error Rates in Quantum Hypothesis Testing, arXiv:0708.4282v1, PDF p. 23, §5.2, conditional rate shifts (43)
Read-back

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

For every finite index type III with decidable equality, let p,q:I→Rp,q:I\to\mathbb Rp,q:I→R be nonnegative functions summing to 111, let C={i:pi>0 and qi>0}C=\{i:p_i>0\ \text{and}\ q_i>0\}C={i:pi​>0 and qi​>0}, and assume P=∑i∈Cpi>0P=\sum_{i\in C}p_i>0P=∑i∈C​pi​>0 and Q=∑i∈Cqi>0Q=\sum_{i\in C}q_i>0Q=∑i∈C​qi​>0. For every natural number kkk, let fkf_kfk​ be a Boolean test on the words x∈Ik+1x\in I^{k+1}x∈Ik+1 having positive probability under at least one of the product laws p(k+1)(x)=∏j=0kpxjp^{(k+1)}(x)=\prod_{j=0}^{k}p_{x_j}p(k+1)(x)=∏j=0k​pxj​​ and q(k+1)(x)=∏j=0kqxjq^{(k+1)}(x)=\prod_{j=0}^{k}q_{x_j}q(k+1)(x)=∏j=0k​qxj​​; assume that fk(x)=falsef_k(x)=\mathrm{false}fk​(x)=false when q(k+1)(x)=0q^{(k+1)}(x)=0q(k+1)(x)=0, and fk(x)=truef_k(x)=\mathrm{true}fk​(x)=true when p(k+1)(x)=0p^{(k+1)}(x)=0p(k+1)(x)=0. Define observable errors αk=∑x:fk(x)=truep(k+1)(x)\alpha_k=\sum_{x:f_k(x)=\mathrm{true}}p^{(k+1)}(x)αk​=∑x:fk​(x)=true​p(k+1)(x) and βk=∑x:fk(x)=falseq(k+1)(x)\beta_k=\sum_{x:f_k(x)=\mathrm{false}}q^{(k+1)}(x)βk​=∑x:fk​(x)=false​q(k+1)(x), with sums over that observable-word domain, and restricted errors

α^k=∑x∈Ck+1fk(x)=true∏j=0kpxjP,β^k=∑x∈Ck+1fk(x)=false∏j=0kqxjQ.\widehat\alpha_k=\sum_{\substack{x\in C^{k+1}\\f_k(x)=\mathrm{true}}} \prod_{j=0}^{k}\frac{p_{x_j}}P, \qquad \widehat\beta_k=\sum_{\substack{x\in C^{k+1}\\f_k(x)=\mathrm{false}}} \prod_{j=0}^{k}\frac{q_{x_j}}Q.αk​=x∈Ck+1fk​(x)=true​∑​j=0∏k​Ppxj​​​,β​k​=x∈Ck+1fk​(x)=false​∑​j=0∏k​Qqxj​​​.

For a real sequence uku_kuk​, “rate vvv” means convergence of −L(uk)/(k+1)-L(u_k)/(k+1)−L(uk​)/(k+1) to the extended real vvv, where L(u)=log⁡uL(u)=\log uL(u)=logu for u>0u>0u>0 and L(u)=−∞L(u)=-\inftyL(u)=−∞ otherwise. For every extended real aaa, α^k\widehat\alpha_kαk​ has rate aaa if and only if αk\alpha_kαk​ has rate −log⁡P+a-\log P+a−logP+a; for every extended real bbb, β^k\widehat\beta_kβ​k​ has rate bbb if and only if βk\beta_kβk​ has rate −log⁡Q+b-\log Q+b−logQ+b. Furthermore, αk\alpha_kαk​ has some extended-real rate if and only if α^k\widehat\alpha_kαk​ does, and βk\beta_kβk​ has some extended-real rate if and only if β^k\widehat\beta_kβ​k​ does. Infinite rates are included, zero errors contribute +∞+\infty+∞ to rate terms, and the word length is always k+1k+1k+1, hence positive.

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