Unnumbered normalization and product preservation
OpenQAlgorithms.Papers.Independent.Intakenext100.P45.mappingProductsFor the spectral-pair map , , both and are normalized probability laws. Their tensor products are orthonormal eigenbases of the tensor-copy states, and preserve the map:
Here are obtained from the product eigenvalues and the overlaps of the product eigenvectors. The empty tensor power is included and gives the singleton law.
import Definitions.Def_QAlgorithms_Papers_Independent_Intakenext100_P45
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.10: Unnumbered normalization and product preservation. -/
theorem mappingProducts (rho sigma : Mat I)
(hrho : IsDensityMatrix rho) (hsigma : IsDensityMatrix sigma)
(D : SpectralPair rho sigma) :
IsLaw (spectralP D) ∧ IsLaw (spectralQ D) ∧
(∀ n : ℕ,
(copies n (D.leftBasis : Mat I))ᴴ * copies n (D.leftBasis : Mat I) = 1 ∧
copies n (D.leftBasis : Mat I) * (copies n (D.leftBasis : Mat I))ᴴ = 1 ∧
(copies n (D.rightBasis : Mat I))ᴴ * copies n (D.rightBasis : Mat I) = 1 ∧
copies n (D.rightBasis : Mat I) * (copies n (D.rightBasis : Mat I))ᴴ = 1 ∧
copies n rho = copies n (D.leftBasis : Mat I) *
Matrix.diagonal (fun x => ((∏ k, D.leftValues (x k) : ℝ) : ℂ)) *
(copies n (D.leftBasis : Mat I))ᴴ ∧
copies n sigma = copies n (D.rightBasis : Mat I) *
Matrix.diagonal (fun x => ((∏ k, D.rightValues (x k) : ℝ) : ℂ)) *
(copies n (D.rightBasis : Mat I))ᴴ) ∧
(∀ n : ℕ, ∀ x y : Fin n → I,
(∏ k, D.leftValues (x k)) *
Complex.normSq (∑ z : Fin n → I,
star (∏ k, (D.leftBasis : Mat I) (z k) (x k)) *
(∏ k, (D.rightBasis : Mat I) (z k) (y k))) =
∏ k, spectralP D (x k,y k)) ∧
(∀ n : ℕ, ∀ x y : Fin n → I,
(∏ k, D.rightValues (y k)) *
Complex.normSq (∑ z : Fin n → I,
star (∏ k, (D.leftBasis : Mat I) (z k) (x k)) *
(∏ k, (D.rightBasis : Mat I) (z k) (y k))) =
∏ k, spectralQ D (x k,y k)) := by sorry
end QAlgorithms.Papers.Independent.Intakenext100.P45
end
Read-back
What the Lean code literally says, in plain math · gpt-6.1-sol
For every finite index type with decidable equality, every positive semidefinite complex matrices indexed by with trace one, and every unitary matrices and nonnegative real families satisfying and , define
Both and are nonnegative and sum to one over . For every natural number , define , indexed by functions , by . Then
and
Furthermore, for every such ,
and
These statements include : all empty products are one, and each matrix is the one-by-one matrix with entry one. The density hypotheses exclude an empty .
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.