Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

Section IV A, Eq.(12) and reference note [45]: at most 2d^2 observable averages

Proved
QAlgorithms.Papers.Independent.Advance50.P17.reconstruction12

by Gohan · Oct 29, 2026 · Mathlib 0df444a (Lean v4.33.1)

prove2quantum

Let E\mathcal EE be CPTP and UUU unitary on a register of dimension d≥1d\ge1d≥1. Choose an orthogonal unitary basis VjV_jVj​, with tr⁡(Vj†Vk)=dδjk\operatorname{tr}(V_j^\dagger V_k)=d\delta_{jk}tr(Vj†​Vk​)=dδjk​, and any density-matrix operator basis ρk\rho_kρk​. Compute ajka_{jk}ajk​ from Vj=∑kajkρkV_j=\sum_k a_{jk}\rho_kVj​=∑k​ajk​ρk​ and set σk=∑jajkUVj†U†\sigma_k=\sum_j a_{jk}UV_j^\dagger U^\daggerσk​=∑j​ajk​UVj†​U†. Then

Fpro(E,U)=d−3∑ktr⁡(σkE(ρk)).F_{\rm pro}(\mathcal E,U)=d^{-3}\sum_k\operatorname{tr}(\sigma_k\mathcal E(\rho_k)).Fpro​(E,U)=d−3k∑​tr(σk​E(ρk​)).

Writing Ak=(σk+σk†)/2A_k=(\sigma_k+\sigma_k^\dagger)/2Ak​=(σk​+σk†​)/2 and Bk=(σk−σk†)/(2i)B_k=(\sigma_k-\sigma_k^\dagger)/(2i)Bk​=(σk​−σk†​)/(2i) gives Hermitian observables with σk=Ak+iBk\sigma_k=A_k+iB_kσk​=Ak​+iBk​. Hence at most 2d22d^22d2 observable averages suffice, or d2d^2d2 when the paired operators are Hermitian. These are counts of distinct averages, not samples. The adjoint corrects the printed formula consistently with (10).

Preamble
import Definitions.Def_QAlgorithms_Papers_Independent_Advance50_P17
Formal statement
noncomputable section

open MeasureTheory Filter

open scoped Matrix ComplexConjugate Topology

namespace QAlgorithms.Papers.Independent.Advance50.P17

open QAlgorithms.Papers.Independent.Information100.P53 (fidelity)

open QAlgorithms.Papers.Independent.Information100.P08 (canonicalPurification)

open QAlgorithms.Papers.Independent.Information100.P75 (hermitianReal)

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

/-- Section IV A, PDF p.6, Eq.(12), and note [45], p.15: at most 2d^2 Hermitian averages.
The adjoint in pairedOperator corrects the same printed omission as Eq.(11). -/
theorem reconstruction12 (E : FiniteProcess I I) (U : Matrix.unitaryGroup I ℂ)
    (V : I × I → Matrix.unitaryGroup I ℂ)
    (hv : ∀ j k, ((V j : Matrix I I ℂ)ᴴ * (V k : Matrix I I ℂ)).trace =
      if j = k then (Fintype.card I : ℂ) else 0)
    (r : Module.Basis (I × I) ℂ (Matrix I I ℂ)) (hr : ∀ k, IsDensityMatrix (r k)) :
    (proFidelity E.val (conjSuperop (U : Matrix I I ℂ)) : ℂ) =
      (Fintype.card I : ℂ)⁻¹ ^ 3 * ∑ k, (pairedOperator r V U k * E.val (r k)).trace ∧
    (∀ k, (hermitianReal (pairedOperator r V U k)).IsHermitian ∧
      (hermitianImag (pairedOperator r V U k)).IsHermitian ∧
      pairedOperator r V U k = hermitianReal (pairedOperator r V U k) +
        Complex.I • hermitianImag (pairedOperator r V U k)) ∧
    Fintype.card ((I × I) × Bool) = 2 * Fintype.card I ^ 2 ∧
    ((∀ k, (pairedOperator r V U k).IsHermitian) →
      proFidelity E.val (conjSuperop (U : Matrix I I ℂ)) =
        (Fintype.card I : ℝ)⁻¹ ^ 3 *
          ∑ k, effectProb (pairedOperator r V U k) (E.val (r k))) ∧
    Fintype.card (I × I) = Fintype.card I ^ 2 := by sorry

end QAlgorithms.Papers.Independent.Advance50.P17

end
Source
Distance measures to compare real and ideal quantum processes, arXiv:quant-ph/0408063v2, PDF p. 6, Section IV A, Eq.(12) and reference note [45]: at most 2d^2 observable averages
Read-back

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

Let III be a finite nonempty type with decidable equality, put d=∣I∣d=|I|d=∣I∣, and let EEE be a channel on its complex matrix space, admitting a finite Kraus representation E(X)=∑jKjXKj†E(X)=\sum_jK_jXK_j^\daggerE(X)=∑j​Kj​XKj†​ with ∑jKj†Kj=1\sum_jK_j^\dagger K_j=1∑j​Kj†​Kj​=1. Let UUU and all matrices VjV_jVj​, indexed by j∈I×Ij\in I\times Ij∈I×I, be unitary, and assume tr⁡(Vj†Vk)=d\operatorname{tr}(V_j^\dagger V_k)=dtr(Vj†​Vk​)=d when j=kj=kj=k and 000 otherwise. Let (rk)k∈I×I(r_k)_{k\in I\times I}(rk​)k∈I×I​ be a complex basis of the matrix space consisting of positive semidefinite trace-one matrices. Using its coordinate functions rk∗r_k^*rk∗​, set

Bk=∑j∈I×Irk∗(Vj) UVj†U†,Hk=12(Bk+Bk†),Kk=12i(Bk−Bk†).B_k=\sum_{j\in I\times I}r_k^*(V_j)\,UV_j^\dagger U^\dagger,\qquad H_k=\tfrac12(B_k+B_k^\dagger),\qquad K_k=\tfrac1{2i}(B_k-B_k^\dagger).Bk​=j∈I×I∑​rk∗​(Vj​)UVj†​U†,Hk​=21​(Bk​+Bk†​),Kk​=2i1​(Bk​−Bk†​).

Define ϕ(i,j)=d−1/2\phi(i,j)=d^{-1/2}ϕ(i,j)=d−1/2 for i=ji=ji=j and 000 otherwise, CG=(id⁡I⊗G)(ϕϕ†)C_G=(\operatorname{id}_I\otimes G)(\phi\phi^\dagger)CG​=(idI​⊗G)(ϕϕ†), with GGG applied blockwise, and Fpro(G,H)=(Re⁡tr⁡CGCHCG)2F_{\mathrm{pro}}(G,H)=(\operatorname{Re}\operatorname{tr}\sqrt{\sqrt{C_G}C_H\sqrt{C_G}})^2Fpro​(G,H)=(RetrCG​​CH​CG​​​)2. Then, writing CU(X)=UXU†\mathcal C_U(X)=UXU^\daggerCU​(X)=UXU†, one has the complex equality

Fpro(E,CU)=d−3∑ktr⁡(BkE(rk));F_{\mathrm{pro}}(E,\mathcal C_U)=d^{-3}\sum_k\operatorname{tr}(B_kE(r_k));Fpro​(E,CU​)=d−3k∑​tr(Bk​E(rk​));

every Hk,KkH_k,K_kHk​,Kk​ is Hermitian and Bk=Hk+iKkB_k=H_k+iK_kBk​=Hk​+iKk​; and ∣(I×I)×{false,true}∣=2d2|(I\times I)\times\{\mathrm{false},\mathrm{true}\}|=2d^2∣(I×I)×{false,true}∣=2d2. Furthermore, if every BkB_kBk​ is Hermitian, then the real equality Fpro(E,CU)=d−3∑kRe⁡tr⁡(BkE(rk))F_{\mathrm{pro}}(E,\mathcal C_U)=d^{-3}\sum_k\operatorname{Re}\operatorname{tr}(B_kE(r_k))Fpro​(E,CU​)=d−3∑k​Retr(Bk​E(rk​)) holds. Finally, ∣I×I∣=d2|I\times I|=d^2∣I×I∣=d2.

Human review
  • Endorsed by Shuze Chen · Oct 29, 2026

    Confirmed by the moderator at approval.

  • Endorsed by Gohan · Oct 29, 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