Documentation

Physicslib4.GNS.CauchySchwarz

Cauchy-Schwarz inequality for positive linear functionals on a *-algebra #

This file states the Cauchy-Schwarz inequality for positive linear functionals on an arbitrary (not necessarily unital, not necessarily normed) *-algebra over , following the AQFT-in-Lean blueprint section 10.1, label lmm:cauchy-schwarz-inequality.

The hypothesis is weaker than Physicslib4.GNS.State: we only require a -linear functional ω : A →ₗ[ℂ] ℂ together with positivity 0 ≤ ω (star a * a) for all a. Continuity and normalisation are not assumed.

Main statements #

Both conclusions of the blueprint lemma are recorded; the primary entry-point is cauchy_schwarz_inequality.

Typeclass setup #

The blueprint says "*-algebra"; the minimal Mathlib shape needed to state the lemma is a complex *-ring, i.e. [NonUnitalNonAssocRing A] [StarRing A] [Module ℂ A]. Continuity of ω and normalisation are not required.

theorem Physicslib4.GNS.omega_star_swap_conj {A : Type u_1} [NonUnitalNonAssocRing A] [StarRing A] [Module A] [IsScalarTower A A] [SMulCommClass A A] [StarModule A] (ω : A →ₗ[] ) ( : ∀ (a : A), 0 ω (star a * a)) (a b : A) :
ω (star a * b) = star (ω (star b * a))

Hermitian symmetry for positive linear functionals on a *-algebra (part 1 of lmm:cauchy-schwarz-inequality).

If ω : A →ₗ[ℂ] ℂ satisfies 0 ≤ ω (star a * a) for every a : A, then ω (star a * b) = star (ω (star b * a)) for all a, b : A, i.e. ω(a*b) = conj(ω(b*a)).

theorem Physicslib4.GNS.cauchy_schwarz_inequality {A : Type u_1} [NonUnitalNonAssocRing A] [StarRing A] [Module A] [IsScalarTower A A] [SMulCommClass A A] [StarModule A] (ω : A →ₗ[] ) ( : ∀ (a : A), 0 ω (star a * a)) (a b : A) :
Complex.normSq (ω (star a * b)) (ω (star a * a)).re * (ω (star b * b)).re

Cauchy-Schwarz inequality for positive linear functionals on a *-algebra (lmm:cauchy-schwarz-inequality, primary entry point).

Let A be a complex *-algebra (here packaged as a NonUnitalNonAssocRing with a compatible StarRing structure and Module ℂ A). If ω : A →ₗ[ℂ] ℂ satisfies 0 ≤ ω (star a * a) for every a : A, then for all a, b : A we have |ω(a*b)|² ≤ ω(a*a) · ω(b*b).

Because ω (star a * a) and ω (star b * b) are non-negative reals (by the positivity hypothesis), the inequality is stated using the real parts of these complex numbers and Complex.normSq for the left-hand side.