Skip to content

Measurable Functions

5.1 Definition

Let (X,F)(X, \mathcal{F}) and (Y,G)(Y, \mathcal{G}) be measurable spaces. A function f:XYf : X \to Y is measurable if f1(G)Ff^{-1}(G) \in \mathcal{F} for every GGG \in \mathcal{G}. When Y=RY = \mathbb{R}, we equip R\mathbb{R} with B(R)\mathcal{B}(\mathbb{R}).

Proposition 5.1. f:XRf : X \to \mathbb{R} is measurable if and only if f1((a,))Ff^{-1}((a, \infty)) \in \mathcal{F} for every aRa \in \mathbb{R}.

Proof. Since (a,)(a, \infty) generates B(R)\mathcal{B}(\mathbb{R}), the σ\sigma-algebra f1(B(R))f^{-1}(\mathcal{B}(\mathbb{R})) equals the σ\sigma-algebra generated by f1((a,))f^{-1}((a, \infty)). \blacksquare

Proposition 5.2. Compositions of measurable functions are measurable.

Proposition 5.3. If f,g:XRf, g : X \to \mathbb{R} are measurable, then f+gf + g, fgfg, f/gf/g (when defined), f|f|, max(f,g)\max(f, g), and min(f,g)\min(f, g) are all measurable.

5.2 Simple Functions

A simple function is a finite linear combination of indicator functions:

s(x)=i=1naiχAi(x)s(x) = \sum_{i=1}^{n} a_i \chi_{A_i}(x)

where aiRa_i \in \mathbb{R} and AiFA_i \in \mathcal{F} are measurable sets.

Theorem 5.4 (Approximation Theorem). For every non-negative measurable function f:X[0,]f : X \to [0, \infty], there exists an increasing sequence of simple functions 0s1s20 \leq s_1 \leq s_2 \leq \cdots converging pointwise to ff.

Proof. For each nn, partition [0,n)[0, n) into n2nn \cdot 2^n subintervals of length 2n2^{-n}. Define

sn(x)={k12nif k12nf(x)<k2n, k=1,,n2nnif f(x)ns_n(x) = \begin{cases} \frac{k-1}{2^n} & \text{if } \frac{k-1}{2^n} \leq f(x) < \frac{k}{2^n},\ k = 1, \ldots, n2^n \\ n & \text{if } f(x) \geq n \end{cases}

Each sns_n is a simple function, snsn+1s_n \leq s_{n+1}, and sn(x)f(x)s_n(x) \to f(x) for every xx. \blacksquare

5.3 Egorov”s Theorem and Lusin’s Theorem

Theorem 5.5 (Egorov’s Theorem). Let (X,F,μ)(X, \mathcal{F}, \mu) be a finite measure space and let fn:XRf_n : X \to \mathbb{R} be measurable functions converging pointwise to ff a.e. Then for every ε>0\varepsilon > 0, there exists AFA \in \mathcal{F} with μ(A)<ε\mu(A) < \varepsilon such that fnff_n \to f uniformly on XAX \setminus A.

Theorem 5.6 (Lusin’s Theorem). Let f:[a,b]Rf : [a, b] \to \mathbb{R} be Lebesgue measurable. Then for every ε>0\varepsilon > 0, there exists a compact set K[a,b]K \subseteq [a, b] with m([a,b]K)<εm([a, b] \setminus K) < \varepsilon such that fKf|_K is continuous.