Fundamental finite-block error inequality
OpenQAlgorithms.Papers.Independent.Gap48.P41.fundamental_error_boundLet be a nonempty finite orthonormal basis, and let be positive semidefinite complex matrices of trace one with . For a positive integer , an arbitrary effect accepts the null state . Write and . All logarithms are natural, with in rates. Write , using support-restricted spectral logarithms inside the trace.
Put and . Spectral powers vanish on the kernel, including at zero and negative orders.
The maximum defining is attained. For every positive , every , and every effect ,
This finite-block inequality provides the quantitative link between the two testing errors.
import Definitions.Def_QAlgorithms_Papers_Independent_Gap48_P41
noncomputable section
namespace QAlgorithms.Papers.Independent.Gap48.P41
open scoped Matrix ComplexOrder BigOperators
open Filter QAlgorithms
open QAlgorithms.Papers.Independent.Information100.P60 (Mat copies)
open QAlgorithms.Papers.Independent.Information100.P71 (supportPower)
open QAlgorithms.Papers.Independent.Information100.P07 (positiveProjection)
open QAlgorithms.Papers.Independent.Dependency50.P28
(tr SupportLE relativeEntropy scalarPower classicalD typeI accept TestSequence
scaled logNonnegativeExtended)
open QAlgorithms.Papers.Independent.Claimed50.P15 (legendre)
open QAlgorithms.Papers.Independent.Information100.P56 (IsLaw)
attribute [local instance] Classical.propDecidable Classical.decEq
variable {I : Type} [Fintype I] [DecidableEq I]
/-- PDF pp.3--4, Theorem 1, (6)--(8): the finite-block fundamental bound.
The maximum defining phi is attained on [0,1], not on all nonnegative orders. -/
theorem fundamental_error_bound [Nonempty I]
(rho sigma : Mat I) (hrho : IsDensityMatrix rho) (hsigma : IsDensityMatrix sigma)
(hsupp : SupportLE rho sigma) (n : ℕ) (hn : 0 < n) (lambda : ℝ)
(A : Mat (Fin n → I)) (hA : IsEffect A) :
IsGreatest ((fun s : ℝ => lambda * s - petzPsi rho sigma s) '' Set.Icc 0 1)
(petzPhi rho sigma lambda) ∧
1 - typeI rho n A ≤ Real.exp (-(n : ℝ) * petzPhi rho sigma lambda) +
Real.exp ((n : ℝ) * lambda) * accept sigma n A := by sorry
end QAlgorithms.Papers.Independent.Gap48.P41
end
Read-back
What the Lean code literally says, in plain math · gpt-6.1-sol
Let be a nonempty finite type with decidable equality, and let be complex matrices indexed by , each positive semidefinite with trace , such that . For a positive semidefinite matrix , write for the matrix obtained by spectral functional calculus from the function that is at and otherwise; in particular, zero eigenvalues remain zero even when . Define
Here the real logarithm is total, with . For every positive integer , every real , and every matrix indexed by such that and are positive semidefinite, belongs to the set and is at least every member of that set, and
The tensor power has entries ; the hypothesis excludes .
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.