Documentation

TauCeti.Analysis.CStarAlgebra.Action.Covariant.Integrated

Vector-valued integrated forms of covariant pairs #

For a compactly supported continuous coefficient function f, the integrated form is the bounded operator v ↦ ∫ g, π (f g) (U g v). Integration is in the Hilbert space, not in the operator-norm space: strong continuity of U does not imply norm continuity. The operator depends linearly on f and has norm at most ∫ g, ‖f g‖. No separability, second countability, or σ-finiteness is needed. For the definition, the norm bound and the covariance identity the measure need only be finite on compact sets; left invariance enters only for multiplicativity under twisted convolution.

Covariance moves a coefficient on the right of an integrated operator inside its integral with the action twist, while a coefficient on the left comes from the pointwise left multiplier, (π ⋊ U)(a • f) = π(a) (π ⋊ U)(f). For scalar multiples of a fixed coefficient, the construction agrees with ContRepresentation.integratedOperatorL1, after pulling the group representation back along MulEquiv.multiplicativeAdditive.

For a left-invariant measure the integrated form is multiplicative for twisted convolution, (π ⋊ U) (f ⋆ g) = (π ⋊ U) f * (π ⋊ U) g, and turns the twisted left translation CStarAction.translate into the group representation, (π ⋊ U)(r • f) = U_r (π ⋊ U)(f). The order of integration is exchanged by Fubini's theorem for compactly supported continuous functions, so σ-finiteness is still not needed. Extension to coefficient-valued L¹ and the involution identity require additional arguments.

The conventions follow Dana Williams, Crossed Products of C-Algebras*, Chapter 2.

The vector integrand of the integrated form is continuous.

The vector integrand vanishes outside the support of the coefficient function.

The integrated form on compactly supported coefficient functions, defined by vector integration rather than operator-valued Bochner integration.

Instances For

    Matrix coefficients can be computed by integrating vector matrix coefficients.

    A coefficient on the right moves inside the integral with the action twist.

    @[simp]

    The integrated form turns left multiplication by a coefficient into the coefficient representation: (π ⋊ U)(a • f) = π(a) (π ⋊ U)(f).

    theorem TauCeti.CStarAction.CovariantRepresentation.integral_coefficient_apply {G : Type u_1} {A : Type u_2} {H : Type u_3} [Group G] [TopologicalSpace G] [NonUnitalCStarAlgebra A] [NormedAddCommGroup H] [InnerProductSpace ℂ H] [CompleteSpace H] {α : CStarAction G A} (P : α.CovariantRepresentation H) {X : Type u_4} [MeasurableSpace X] {ν : MeasureTheory.Measure X} {F : X → A} (hF : MeasureTheory.Integrable F ν) (w : H) :
    ∫ (x : X), (P.coefficient (F x)) w ∂ν = (P.coefficient (∫ (x : X), F x ∂ν)) w

    Vector evaluation of the coefficient representation commutes with Bochner integration.

    @[simp]

    The integrated form of a covariant pair is multiplicative for twisted convolution with respect to a left-invariant measure: (π ⋊ U) (f ⋆ g) = (π ⋊ U) f * (π ⋊ U) g.

    @[simp]

    The integrated form turns the twisted left translation into the group representation: (π ⋊ U)(r • f) = U_r (π ⋊ U)(f).

    @[reducible]

    Use the original measurable structure on the additive tag for the scalar comparison. Reducibility keeps the measure and its L¹ space unchanged.

    Instances For

      On scalar multiples of a fixed coefficient, the integrated form factors through the existing scalar L¹ integrated form. The tagging equivalence permits an arbitrary multiplicative group. In particular, for the canonical scalar coefficient representation this recovers precisely the scalar integrated operator.