Convexity of the log quasi-entropy moment
ProvedQAlgorithms.Papers.Independent.Advance50.P07.psi_convexLet and act on a nonzero finite-dimensional Hilbert space. Assume the trace below is strictly positive for every , as in the preceding paragraph of VI.B.2. With spectral powers supported on the operators' supports, define
The real-valued function is convex on . No trace normalization is required. This supplies the convex log-moment object used in the paper's entropy analysis.
import Definitions.Def_QAlgorithms_Papers_Independent_Advance50_P07
noncomputable section
open scoped Matrix ComplexConjugate ComplexOrder Topology
open Filter
namespace QAlgorithms.Papers.Independent.Advance50.P07
attribute [local instance] Classical.propDecidable Classical.decEq
open QAlgorithms.Papers.Independent.Information100.P71 (supportPower SupportIncluded)
open QAlgorithms.Papers.Independent.Dependency50.P09 (relativeD petzD)
open QAlgorithms.Papers.Independent.Dependency50.P35 (maxRelative Density elog)
open QAlgorithms.Papers.Independent.Information100.P63 (Mat)
open QAlgorithms.Papers.Independent.Dependency50.P19 (marginalRA marginalRB)
open QAlgorithms.Papers.Independent.Support50.P21 (PositiveTensorBasis)
/-- Lemma 8, PDF pp.6–7. The preceding paragraph's nonorthogonality
assumption is strict positivity of every moment on alpha>0. -/
theorem psi_convex {I : Type} [Fintype I] [DecidableEq I] [Nonempty I]
(rho sigma P : Mat I) (hrho : rho.PosSemidef) (hsigma : sigma.PosSemidef)
(hP : IsEffect P) (hfinite : ∀ alpha : ℝ, 0 < alpha → 0 < quasiMoment rho sigma P alpha) :
ConvexOn ℝ (Set.Ioi (0 : ℝ)) (quasiLogMoment rho sigma P) := by sorry
end QAlgorithms.Papers.Independent.Advance50.P07
end
Read-back
What the Lean code literally says, in plain math · gpt-6.1-sol
Let be any finite, nonempty type with decidable equality, and let be complex matrices indexed by . Assume that and are positive semidefinite and that both and are positive semidefinite. For a Hermitian matrix , write for the matrix obtained by spectral functional calculus from the function that is at the eigenvalue and is at every nonzero eigenvalue ; in particular, is its support projection, including when is singular. Define
Assume for every real . Then is convex on : for all and all real with ,
There is no trace normalization assumption on or ; the strict positivity hypothesis excludes any choice for which one of these moments vanishes.
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.