TheoremBase

A function into a Euclidean space is measurable exactly when its components are; the components of the concatenation are coordinates of the two projections, the inverse is the pairing of two coordinate selections, and e has the identity as its only component.

Proof

Throughout, for p∈Np\in\mathbb{N} the pair (Rp,B(Rp))(\mathbb{R}^{p},\mathcal{B}(\mathbb{R}^{p})) is a measurable space and a function into Rp\mathbb{R}^{p} is measurable when it is measurable for B(Rp)\mathcal{B}(\mathbb{R}^{p}) (Measurable Map, and the Set of Measurable Maps (iv)); the criterion used for all three clauses is

u:W→Rp is measurable  ⟺  each component function w↦u(w)k (k∈[p]) is measurable,u:W\to\mathbb{R}^{p}\text{ is measurable}\iff\text{each component function }w\mapsto u(w)_{k}\ (k\in[p])\text{ is measurable},

for every measurable space (W,K)(W,\mathcal{K}), by Vector- and Matrix-Valued Measurable Maps: Components, Entries and Continuous Images (i) (components as in Component Maps and Entry Maps (i)). For k∈[p]k\in[p] let πk:Rp→R\pi_{k}:\mathbb{R}^{p}\to\mathbb{R}, πk(z)=zk\pi_{k}(z)=z_{k}, be the kk-th component function of the identity of Rp\mathbb{R}^{p}. The identity is B(Rp)/B(Rp)\mathcal{B}(\mathbb{R}^{p})/\mathcal{B}(\mathbb{R}^{p})-measurable (Measurable Maps: Generators, Composition, Continuous Maps, Restrictions, and Criteria for Real- and [0,∞]-Valued Maps (iii)), so by the criterion every coordinate function πk\pi_{k} is B(Rp)/B(R)\mathcal{B}(\mathbb{R}^{p})/\mathcal{B}(\mathbb{R})-measurable. Write P=B(Rm)⊗B(Rn)\mathcal{P}=\mathcal{B}(\mathbb{R}^{m})\otimes\mathcal{B}(\mathbb{R}^{n}), and let pr1:Rm×Rn→Rm\mathrm{pr}_{1}:\mathbb{R}^{m}\times\mathbb{R}^{n}\to\mathbb{R}^{m} and pr2:Rm×Rn→Rn\mathrm{pr}_{2}:\mathbb{R}^{m}\times\mathbb{R}^{n}\to\mathbb{R}^{n} be the projections, which are P/B(Rm)\mathcal{P}/\mathcal{B}(\mathbb{R}^{m})- and P/B(Rn)\mathcal{P}/\mathcal{B}(\mathbb{R}^{n})-measurable by The Product σ-Algebra: Rectangles, Generators, Sections, Projections, Pairings and the Swap (vi).

Concatenation. Let k∈[m+n]k\in[m+n]. By Concatenation Identifies a Product of Euclidean Spaces with a Euclidean Space (i), either k∈[m]k\in[m], or k=m+jk=m+j for some j∈[n]j\in[n]. By Concatenation Identifies a Product of Euclidean Spaces with a Euclidean Space (ii), in the first case

ι(ξ,η)k=ξk=πk(pr1(ξ,η)),\iota(\xi,\eta)_{k}=\xi_{k}=\pi_{k}\bigl(\mathrm{pr}_{1}(\xi,\eta)\bigr),

so the kk-th component function of ι\iota is πk∘pr1\pi_{k}\circ\mathrm{pr}_{1}; in the second case

ι(ξ,η)m+j=ηj=πj(pr2(ξ,η)),\iota(\xi,\eta)_{m+j}=\eta_{j}=\pi_{j}\bigl(\mathrm{pr}_{2}(\xi,\eta)\bigr),

so it is πj∘pr2\pi_{j}\circ\mathrm{pr}_{2}. Either way it is P/B(R)\mathcal{P}/\mathcal{B}(\mathbb{R})-measurable as a composite of measurable functions (Measurable Maps: Generators, Composition, Continuous Maps, Restrictions, and Criteria for Real- and [0,∞]-Valued Maps (ii)). By the criterion, ι\iota is P/B(Rm+n)\mathcal{P}/\mathcal{B}(\mathbb{R}^{m+n})-measurable. This proves (i).

The inverse. Write κ=ι−1\kappa=\iota^{-1}, and let a=pr1∘κ:Rm+n→Rma=\mathrm{pr}_{1}\circ\kappa:\mathbb{R}^{m+n}\to\mathbb{R}^{m} and b=pr2∘κ:Rm+n→Rnb=\mathrm{pr}_{2}\circ\kappa:\mathbb{R}^{m+n}\to\mathbb{R}^{n}, so that κ(w)=(a(w),b(w))\kappa(w)=(a(w),b(w)) for w∈Rm+nw\in\mathbb{R}^{m+n}. By Concatenation Identifies a Product of Euclidean Spaces with a Euclidean Space (iii),

a(w)k=wk=πk(w)(k∈[m]),b(w)j=wm+j=πm+j(w)(j∈[n]),a(w)_{k}=w_{k}=\pi_{k}(w)\quad(k\in[m]),\qquad b(w)_{j}=w_{m+j}=\pi_{m+j}(w)\quad(j\in[n]),

with πk\pi_{k} and πm+j\pi_{m+j} the coordinate functions of Rm+n\mathbb{R}^{m+n}. So the component functions of aa and of bb are coordinate functions, which are measurable, and by the criterion aa is B(Rm+n)/B(Rm)\mathcal{B}(\mathbb{R}^{m+n})/\mathcal{B}(\mathbb{R}^{m})-measurable and bb is B(Rm+n)/B(Rn)\mathcal{B}(\mathbb{R}^{m+n})/\mathcal{B}(\mathbb{R}^{n})-measurable. By The Product σ-Algebra: Rectangles, Generators, Sections, Projections, Pairings and the Swap (vii), κ\kappa is B(Rm+n)/P\mathcal{B}(\mathbb{R}^{m+n})/\mathcal{P}-measurable.

Now let E⊆Rm×RnE\subseteq\mathbb{R}^{m}\times\mathbb{R}^{n}. Then

ι[E]=κ−1(E),E=ι−1(ι[E]),\iota[E]=\kappa^{-1}(E),\qquad E=\iota^{-1}\bigl(\iota[E]\bigr),

where ι−1(⋅)\iota^{-1}(\cdot) and κ−1(⋅)\kappa^{-1}(\cdot) denote preimages (Images and Preimages under a Bijection and Its Inverse (i) and Images and Preimages under a Bijection and Its Inverse (ii), applied to the bijection ι\iota with inverse κ\kappa). If E∈PE\in\mathcal{P}, then ι[E]=κ−1(E)∈B(Rm+n)\iota[E]=\kappa^{-1}(E)\in\mathcal{B}(\mathbb{R}^{m+n}) by measurability of κ\kappa; if ι[E]∈B(Rm+n)\iota[E]\in\mathcal{B}(\mathbb{R}^{m+n}), then E=ι−1(ι[E])∈PE=\iota^{-1}(\iota[E])\in\mathcal{P} by (i). This proves (ii).

The real line and R1\mathbb{R}^{1}. Here [1]={1}[1]=\{1\} (Intervals of Natural Numbers: Initial Segments, Adding One Element, Splitting and Shifting (i)), so an element of R1\mathbb{R}^{1} is determined by its first component (Euclidean Space Rn\mathbb{R}^n: n-Tuples of Real Numbers with Their Componentwise Operations (iii)). The function τ\tau exists as in Sets and Maps: Ordinary Notation (v), and for t∈Rt\in\mathbb{R} and x∈R1x\in\mathbb{R}^{1},

π1(τ(t))=τ(t)1=t,τ(π1(x))=x,\pi_{1}(\tau(t))=\tau(t)_{1}=t,\qquad \tau(\pi_{1}(x))=x,

the second because both sides are 11-tuples with first component x1x_{1}. So τ\tau is a bijection with inverse π1:x↦x1\pi_{1}:x\mapsto x_{1}, by Basic Properties of Functions: Equality, Composition, Identity, Inverse and Restriction (vii). The coordinate function π1\pi_{1} is B(R1)/B(R)\mathcal{B}(\mathbb{R}^{1})/\mathcal{B}(\mathbb{R})-measurable, as shown above. The only component function of τ\tau is π1∘τ=idR\pi_{1}\circ\tau=\mathrm{id}_{\mathbb{R}}, which is B(R)/B(R)\mathcal{B}(\mathbb{R})/\mathcal{B}(\mathbb{R})-measurable (Measurable Maps: Generators, Composition, Continuous Maps, Restrictions, and Criteria for Real- and [0,∞]-Valued Maps (iii)); by the criterion, τ\tau is B(R)/B(R1)\mathcal{B}(\mathbb{R})/\mathcal{B}(\mathbb{R}^{1})-measurable.

Finally let B⊆RB\subseteq\mathbb{R}. By Images and Preimages under a Bijection and Its Inverse (i) and Images and Preimages under a Bijection and Its Inverse (ii), applied to the bijection τ\tau with inverse π1\pi_{1},

τ[B]=π1−1(B),B=τ−1(τ[B]).\tau[B]=\pi_{1}^{-1}(B),\qquad B=\tau^{-1}\bigl(\tau[B]\bigr).

If B∈B(R)B\in\mathcal{B}(\mathbb{R}), then τ[B]=π1−1(B)∈B(R1)\tau[B]=\pi_{1}^{-1}(B)\in\mathcal{B}(\mathbb{R}^{1}) by measurability of π1\pi_{1}; if τ[B]∈B(R1)\tau[B]\in\mathcal{B}(\mathbb{R}^{1}), then B=τ−1(τ[B])∈B(R)B=\tau^{-1}(\tau[B])\in\mathcal{B}(\mathbb{R}) by measurability of τ\tau. This proves (iii).

Citations

Loading…

Dependencies

Uses0

Loading…

Comments

Log in to comment.

Loading…