Integral of vector-valued function against vector measure #
We extend the definition of the Bochner integral (of vector-valued function against ℝ≥0∞-valued
measure) to vector measures through a bilinear pairing.
Let E, F be normed vector spaces, and G be a Banach space (complete normed vector space).
We fix a continuous linear pairing B : E →L[ℝ] F →L[ℝ] G and an F-valued vector measure μ
on a measurable space X.
For an integrable function f : X → E with respect to the total variation of the vector measure
on X informally written μ ∘ B.flip, we define the G-valued integral, which is informally
written ∫ B (f x) ∂μ x.
Such integral is defined through the general setting setToFun which sends a set function to the
integral of integrable functions, see the file
Mathlib/MeasureTheory/Integral/SetToL1.lean.
Main definitions #
The integral against vector measures is defined through the extension process described in the file
Mathlib/MeasureTheory/Integral/SetToL1.lean, which follows these steps:
Define the integral of the indicator of a set. This is
cbmApplyMeasure B μ s x = B x (μ s).cbmApplyMeasure B μis shown to be linear in the valuexandDominatedFinMeasAdditive(defined in the fileMathlib/MeasureTheory/Integral/SetToL1.lean) with respect to the sets.Define the integral on integrable functions
fassetToFun (...) f.
Notations #
∫ᵛ x, f x ∂[B; μ]: theG-valued integral of anE-valued functionfagainst theF-valued vector measureμpaired throughB.∫ᵛ x, f x ∂•μ: the special case wherefis a real-valued function andμis anF-valued vector measure, with the pairing being the scalar multiplication byℝ.∫ᵛ x, f x ∂<•μ: the special case wherefis anE-valued function andμis a signed measure, with the pairing being the flip of scalar multiplication.∫ᵛ x in s, f x ∂[B; μ]: theG-valued integral of anE-valued functionfagainst theF-valued vector measureμpaired throughB, on the sets.∫ᵛ x in s, f x ∂•μ: the special case wherefis a real-valued function andμis anF-valued vector measure, with the pairing being the scalar multiplication byℝ.∫ᵛ x in s, f x ∂<•μ: the special case wherefis anE-valued function andμis a signed measure, with the pairing being the flip of scalar multiplication.
Note #
Let μ be a vector measure and B be a continuous linear pairing.
We often consider integrable functions with respect to the total variation of
μ.transpose B = μ.mapRange B.flip.toAddMonoidHom B.flip.continuous, which is the reference
measure for the pairing integral.
When f is not integrable with respect to (μ.transpose B).variation, the value of
μ.integral B f is set to 0. This is an analogous convention to the Bochner integral. However,
there are cases where a natural definition of the integral as an unconditional sum exists, but f
is not integrable in this sense: Let μ be the L∞(ℕ)-valued measure on ℕ defined by extending
{n} ↦ (0,0,..., 1/(n+1),0,0,...) and B be the trivial coupling (the scalar multiplication by
ℝ). The total variation is ∑ n, 1/(n+1) = ∞, but the sum of (0,...,0,1/n,0,...) in L∞(ℕ) is
unconditionally convergent.
The composition of the vector measure with the linear pairing, giving the reference vector measure.
Equations
- μ.transpose B = μ.mapRange (↑B.flip).toAddMonoidHom ⋯
Instances For
Given a set s, return the continuous linear map fun x : E ↦ B x (μ s) (actually defined
using transpose through mapRange), where the B is a G-valued bilinear form on E × F and
μ is an F-valued vector measure. The extension of that set function through setToFun gives the
pairing integral of E-valued integrable functions.
Equations
- MeasureTheory.cbmApplyMeasure μ B s = { toFun := fun (x : E) => (↑(μ.transpose B) s) x, map_add' := ⋯, map_smul' := ⋯, cont := ⋯ }
Instances For
Control of the variation of the vector measure which appears in the integral of scalar functions with respect to a vector measure.
Control of the variation of the vector measure which appears in the integral of a vector function with respect to a signed measure.
f : X → E is said to be integrable with respect to μ and B if it is integrable with
respect to (μ.transpose B).variation.
Equations
- μ.Integrable f B = MeasureTheory.Integrable f (μ.transpose B).variation
Instances For
f : X → E is said to be integrable with respect to μ and B on s if it is integrable with
respect to the vector measure μ.restrict s. When s is measurable, this is equivalent to
integrability with respect to (μ.transpose B).variation.restrict s.
Equations
- μ.IntegrableOn f B s = (μ.restrict s).Integrable f B
Instances For
The G-valued integral of E-valued function and the F-valued vector measure μ with linear
paring B : E →L[ℝ] F →L[ℝ] G . This is set to be 0 if G is not complete or if f is not
integrable with respect to (μ.transpose B).variation. Notation ∫ᵛ x, f x ∂[B; μ].
When μ is G-valued, to get the integral in G of a real-valued function, take
B = ContinousLinearMap.lsmul ℝ ℝ. Notation ∫ᵛ x, f x ∂•μ.
When μ is a signed measure, to get the integral in G of a G-valued function, take
B = (ContinousLinearMap.lsmul ℝ ℝ).flip. Notation ∫ᵛ x, f x ∂<•μ.
Instances For
The G-valued integral of E-valued function and the F-valued vector measure μ with linear
paring B : E →L[ℝ] F →L[ℝ] G . This is set to be 0 if G is not complete or if f is not
integrable with respect to (μ.transpose B).variation. Notation ∫ᵛ x, f x ∂[B; μ].
When μ is G-valued, to get the integral in G of a real-valued function, take
B = ContinousLinearMap.lsmul ℝ ℝ. Notation ∫ᵛ x, f x ∂•μ.
When μ is a signed measure, to get the integral in G of a G-valued function, take
B = (ContinousLinearMap.lsmul ℝ ℝ).flip. Notation ∫ᵛ x, f x ∂<•μ.
Equations
- One or more equations did not get rendered due to their size.
Instances For
The G-valued integral of E-valued function and the F-valued vector measure μ with linear
paring B : E →L[ℝ] F →L[ℝ] G . This is set to be 0 if G is not complete or if f is not
integrable with respect to (μ.transpose B).variation. Notation ∫ᵛ x, f x ∂[B; μ].
When μ is G-valued, to get the integral in G of a real-valued function, take
B = ContinousLinearMap.lsmul ℝ ℝ. Notation ∫ᵛ x, f x ∂•μ.
When μ is a signed measure, to get the integral in G of a G-valued function, take
B = (ContinousLinearMap.lsmul ℝ ℝ).flip. Notation ∫ᵛ x, f x ∂<•μ.
Equations
- One or more equations did not get rendered due to their size.
Instances For
Eta-expanded form of MeasureTheory.VectorMeasure.Integrable.add
Eta-expanded form of MeasureTheory.VectorMeasure.Integrable.neg
Eta-expanded form of MeasureTheory.VectorMeasure.Integrable.sub
Eta-expanded form of MeasureTheory.VectorMeasure.Integrable.smul
If F i → f in L1, then ∫ᵛ x, F i x ∂[B; μ] → ∫ᵛ x, f x ∂[B; μ].
If F i → f in L1, then ∫ᵛ x, F i x ∂[B; μ] → ∫ᵛ x, f x ∂[B; μ].
Lebesgue dominated convergence theorem provides sufficient conditions under which almost
everywhere convergence of a sequence of functions implies the convergence of their integrals.
We could weaken the condition bound_integrable to require
HasFiniteIntegral bound (μ.transpose B).variation instead (i.e. not requiring that bound is
measurable), but in all applications proving integrability is easier.
Lebesgue dominated convergence theorem for filters with a countable basis
Lebesgue dominated convergence theorem for series.
Corollary of the Lebesgue dominated convergence theorem: If a sequence of functions F n is
(eventually) uniformly bounded by a constant and converges (eventually) pointwise to a
function f, then the integrals of F n with respect to a vector measure μ with finite
variation converge to the integral of f.