Documentation

CsdLean4.Mathlib.Analysis.InnerProductSpace.Gleason.Core

Gleason's theorem, finite-dimensional: the core lemma discharged, the theorem proved #

Category: 1-Mathlib (CSD-free; staged for upstream). The end of the staged plan of specs/gleason-feasibility.md (BACKLOG #57): Gleason/General.lean proved the core lemma (frameFunction_regular_sphere, stage 57(e)) and Gleason/Reduction.lean proved the theorem from it. This file joins the two.

What is and is not claimed. This is the finite-dimensional theorem (ℂᴺ, N ≥ 3) for probability assignments on orthogonal projections. The infinite-dimensional theorem, dimension 2 (where the statement fails) and the POVM generalisation (Busch's, LF2/EffectGleason.lean, proved separately) are outside it. Neither of the two theorems is derived from the other: the proofs share the Jordan–von Neumann engine (Gleason/Polarization.lean) and the descent (Gleason/Descent.lean), nothing else.

Source #

Gleason 1957, J. Math. Mech. 6, 885 (Theorem 2.8, the nonnegative core lemma; Theorem 3.1, the theorem); Cooke, Keane, Moran 1985, Math. Proc. Cambridge Philos. Soc. 98, 117 (the elementary proof of the core lemma followed in Gleason/Sphere.lean through Gleason/General.lean).

The core lemma, in the form Reduction.lean consumes.

theorem Gleason.ProjectionPackage.gleason_representation {N : ℕ} (OP : ProjectionPackage N) (hN : 3 ≤ N) :
∃! ρ : Matrix (Fin N) (Fin N) ℂ, ρ.PosSemidef ∧ ρ.trace = 1 ∧ ∀ (P : Matrix (Fin N) (Fin N) ℂ), IsStarProjection P → OP.p P = (ρ * P).trace.re

★★ Gleason's theorem for ℂᴺ, N ≥ 3. For every projection package OP there is a unique density matrix ρ with OP.p P = Re Tr(ρ P) for every orthogonal projection P.