Example 2.7
ProvedQAlgorithms.Papers.Independent.Support50.P12.example_2_7Let be positive semidefinite operators on a nonzero finite-dimensional complex Hilbert space. Let be continuous on with signed extended recession slope . Write for the paper's quantum divergence, , and for the support projection. All powers vanish on the kernel, including exponent zero. Traces are representation traces.
Put for , and for , . Then
For , , the Rényi entropy is with the paper's extended logarithm. Also is on supported pairs and otherwise. These are Petz, rather than sandwiched, formulas, without trace normalization.
import Definitions.Def_QAlgorithms_Papers_Independent_Support50_P12
noncomputable section
namespace QAlgorithms.Papers.Independent.Support50.P12
open Filter MeasureTheory Set
open QAlgorithms.Papers.Independent.Information100.P60 (Mat copies)
open QAlgorithms.Papers.Independent.Information100.P30 (MatrixMap)
open QAlgorithms.Papers.Independent.Dependency50.P28 (tr LE logNonnegativeExtended pinch)
open QAlgorithms.Papers.Independent.Dependency50.P21 (channelAdjoint spectralFunction petz relativeEntropy)
open QAlgorithms.Papers.Independent.Frontier18.P17 (hs)
open QAlgorithms.Papers.Independent.Next46.P23 (Supports)
open QAlgorithms.Papers.Independent.Support21.P03 (ExtendedConvex)
open QAlgorithms (spectralProjector)
open QAlgorithms.Learning (IsDistribution)
open QAlgorithms.Papers.Independent.Information100.P04 (absOp)
open scoped BigOperators Matrix ComplexConjugate ComplexOrder ENNReal Topology Kronecker
attribute [local instance] Classical.propDecidable Classical.decEq
open QAlgorithms.Papers.Independent.Information100.P71 (supportPower)
variable {I J K : Type} [Fintype I] [DecidableEq I] [Fintype J] [DecidableEq J]
[Fintype K] [DecidableEq K]
/-- Example 2.7, PDF pp.6–7: support powers, unnormalized Petz Rényi entropy,
and x log x. The alpha=1 identity is retained. -/
theorem example_2_7 [Nonempty I] (A B : Mat I) (hA : A.PosSemidef) (hB : B.PosSemidef) :
(∀ a : ℝ, 0 ≤ a →
div (supportScalarPower a) A B = (tr (supportPower A a * supportPower B (1 - a)) : EReal) +
recessionSlope (supportScalarPower a) * (tr (A * (1 - supp B)) : EReal)) ∧
(∀ a ∈ Ico (0 : ℝ) 1,
div (supportScalarPower a) A B = (tr (supportPower A a * supportPower B (1 - a)) : EReal)) ∧
(∀ a : ℝ, 1 < a → div (supportScalarPower a) A B =
if Supports A B then (tr (supportPower A a * supportPower B (1 - a)) : EReal) else ⊤) ∧
(∀ a : ℝ, 0 ≤ a → a ≠ 1 →
petzRenyi a A B = (((a - 1)⁻¹ : ℝ) : EReal) *
(if div (supportScalarPower a) A B = ⊤ then ⊤ else logNonnegativeExtended (div (supportScalarPower a) A B).toReal)) ∧
div (fun x => x * Real.log x) A B = relativeEntropy A B := by sorry
end QAlgorithms.Papers.Independent.Support50.P12
end
Read-back
What the Lean code literally says, in plain math · gpt-6.1-sol
Let be a finite, nonempty index type with decidable equality, and let be positive semidefinite complex matrices indexed by . Write , , and ; thus zero eigenvalues contribute zero even when or . Write for . For positive semidefinite , let
where the spectral sums use distinct eigenvalues and their orthogonal eigenspace projections, division by zero is defined as zero, and the second branch uses the formal total limit-selection operation in the extended real line. Put when and otherwise, and let be the extended-real limit-selection value of as . Define for and otherwise, and
Then all of the following hold:
Here extended-real values are used throughout these equalities, and sends either infinity to zero. Finally,
No trace normalization or invertibility is assumed; either matrix may be zero.
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.