Skip to main content
\(\newcommand{\id}{\mathrm{id}}\) \( \newcommand{\Span}{\mathrm{span}}\) \( \newcommand{\kernel}{\mathrm{null}\,}\) \( \newcommand{\range}{\mathrm{range}\,}\) \( \newcommand{\RealPart}{\mathrm{Re}}\) \( \newcommand{\ImaginaryPart}{\mathrm{Im}}\) \( \newcommand{\Argument}{\mathrm{Arg}}\) \( \newcommand{\norm}[1]{\| #1 \|}\) \( \newcommand{\inner}[2]{\langle #1, #2 \rangle}\) \( \newcommand{\Span}{\mathrm{span}}\)
Mathematics LibreTexts

8.3.E: Problems on Measurable Functions in \((S, \mathcal{M}, m)\)

  • Page ID
  • \( \newcommand{\vecs}[1]{\overset { \scriptstyle \rightharpoonup} {\mathbf{#1}} } \) \( \newcommand{\vecd}[1]{\overset{-\!-\!\rightharpoonup}{\vphantom{a}\smash {#1}}} \)\(\newcommand{\id}{\mathrm{id}}\) \( \newcommand{\Span}{\mathrm{span}}\) \( \newcommand{\kernel}{\mathrm{null}\,}\) \( \newcommand{\range}{\mathrm{range}\,}\) \( \newcommand{\RealPart}{\mathrm{Re}}\) \( \newcommand{\ImaginaryPart}{\mathrm{Im}}\) \( \newcommand{\Argument}{\mathrm{Arg}}\) \( \newcommand{\norm}[1]{\| #1 \|}\) \( \newcommand{\inner}[2]{\langle #1, #2 \rangle}\) \( \newcommand{\Span}{\mathrm{span}}\) \(\newcommand{\id}{\mathrm{id}}\) \( \newcommand{\Span}{\mathrm{span}}\) \( \newcommand{\kernel}{\mathrm{null}\,}\) \( \newcommand{\range}{\mathrm{range}\,}\) \( \newcommand{\RealPart}{\mathrm{Re}}\) \( \newcommand{\ImaginaryPart}{\mathrm{Im}}\) \( \newcommand{\Argument}{\mathrm{Arg}}\) \( \newcommand{\norm}[1]{\| #1 \|}\) \( \newcommand{\inner}[2]{\langle #1, #2 \rangle}\) \( \newcommand{\Span}{\mathrm{span}}\)

    Exercise \(\PageIndex{1}\)

    Fill in all proof details in Corollaries 1 to 4.

    Exercise \(\PageIndex{1'}\)

    Verify Notes 3 and 4.

    Exercise \(\PageIndex{2}\)

    Prove Theorems 1 and 2 in §1 and Theorem 2 in §2, for almost measurable functions.

    Exercise \(\PageIndex{3}\)

    Prove Note 2.
    [Hint: If \(f: S \rightarrow E^{*}\) is \(\mathcal{M}\)-measurable on \(B=A-Q(m Q=0, Q \subseteq A),\) then \(A=B \cup Q\) and
    \left(\forall a \in E^{*}\right) \quad A(f>a)=B(f>a) \cup Q(f>a) .
    Here \(B(f>a) \in \mathcal{M}\) by Theorem 1 in §2, and \(Q(f>a) \in \mathcal{M}\) if \(m\) is complete. For \(\left.f: S \rightarrow E^{n}\left(C^{n}\right), \text { use Theorem } 2 \text { of } §1 .\right]\)

    Exercise \(\PageIndex{4}\)

    *4. Show that if \(m\) is complete and \(f: S \rightarrow\left(T, \rho^{\prime}\right)\) is \(m\)-measurable on \(A\) with \(f[A]\) separable in \(T,\) then \(f\) is \(\mathcal{M}\) -measurable on \(A .\)
    [Hint: Use Problem \(13 \text { in } §2 .]\)

    Exercise \(\PageIndex{5}\)

    *5. Prove Theorem 1 for \(f: S \rightarrow\left(T, \rho^{\prime}\right),\) assuming that \(f[A]\) is separable in \(T .\)

    Exercise \(\PageIndex{6}\)

    Given \(f_{n} \rightarrow f(\text { a.e. })\) on \(A,\) prove that \(f_{n} \rightarrow g(\text { a.e. })\) on \(A\) iff \(f=g(\text { a.e. })\) on \(A .\)

    Exercise \(\PageIndex{7}\)

    Given \(A \in \mathcal{M}\) in \((S, \mathcal{M}, m),\) let \(m_{A}\) be the restriction of \(m\) to
    \mathcal{M}_{A}=\{X \in \mathcal{M} | X \subseteq A\} .
    Prove that
    (i) \(\left.\left(A, \mathcal{M}_{A}, m_{A}\right) \text { is a measure space (called a subspace of }(S, \mathcal{M}, m)\right)\);
    (ii) if \(m\) is complete, topological, \(\sigma\)-finite or (strongly) regular, so is \(m_{A}\).

    Exercise \(\PageIndex{8}\)

    (i) Show that if \(D \subseteq K \subseteq\left(T, \rho^{\prime}\right),\) then the closure of \(D\) in the subspace \(\left(K, \rho^{\prime}\right)\) is \(K \cap \bar{D},\) where \(\bar{D}\) is the closure of \(D\) in \(\left(T, \rho^{\prime}\right) .\)
    [Hint: Use Problem \(11 \text { in Chapter } 3, §16 .]\)
    (ii) Prove that if \(B \subseteq K\) and if \(B\) is separable in \(\left(T, \rho^{\prime}\right),\) it is so in \(\left(K, \rho^{\prime}\right) .\)
    [Hint: Use Problem 7 from \(\xi 1\).]

    Exercise \(\PageIndex{9}\)

    *9. Fill in all proof details in Lemma 4.

    Exercise \(\PageIndex{10}\)

    Simplify the proof of Theorem 2 for the case \(m A<\infty .\)
    [Outline: (i) First, let \(f\) be elementary, with \(f=a_{i}\) on \(A_{i} \in \mathcal{M}, A=\cup_{i} A_{i}\) (disjoint), \(\sum m A_{i}=m A<\infty\).
    Given \(\varepsilon>0\)
    (\exists n) \quad m A-\sum_{i=1}^{n} m A_{i}<\frac{1}{2} \varepsilon .
    Each \(A_{i}\) has a closed subset \(F_{i} \in \mathcal{M}\) with \(m\left(A_{i}-F_{i}\right)<\varepsilon / 2 n .\) (Why?) Now use Problem 17 in Chapter 4, §8, and set \(F=\bigcup_{i=1}^{n} F_{i} .\)
    (ii) If \(f\) is \(\mathcal{M}\) -measurable on \(H=A-Q, m Q=0,\) then by Theorem 3 in \(\xi 1,\)
    \(f_{n} \rightarrow f\) (uniformly) on \(H\) for some elementary maps \(f_{n} .\) By \((i),\) each \(f_{n}\) is relatively continuous on a closed \(\mathcal{M}\)-set \(F_{n} \subseteq H,\) with \(m H-m F_{n}<\varepsilon / 2^{n} ;\) so all \(f_{n}\) are relatively continuous on \(F=\bigcap_{n=1}^{\infty} F_{n} .\) Show that \(F\) is the required set.

    Exercise \(\PageIndex{11}\)

    Given \(f_{n}: S \rightarrow\left(T, \rho^{\prime}\right), n=1,2, \ldots,\) we say that
    (i) \(f_{n} \rightarrow f\) almost uniformly on \(A \subseteq S\) iff
    (\forall \delta>0)(\exists D \in \mathcal{M} | m D<\delta) \quad f_{n} \rightarrow f(\text {uniformly}) \text { on } A-D ;
    (ii) \(f_{n} \rightarrow f\) in measure on \(A\) iff
    \begin{aligned}(\forall \delta, \sigma>0)(\exists k)(\forall n>k)\left(\exists D_{n} \in \mathcal{M} | m D_{n}<\delta\right) \\ \rho^{\prime}\left(f, f_{n}\right)<\sigma \text { on } A-D_{n} . \end{aligned}
    Prove the following.
    (a) \(f_{n} \rightarrow f\) (uniformly) implies \(f_{n} \rightarrow f\) (almost uniformly), and the latter implies both \(f_{n} \rightarrow f\left(\text { in measure) and } f_{n} \rightarrow f(a . e .) .\right.\)
    (b) Given \(f_{n} \rightarrow f\) (almost uniformly), we have \(f_{n} \rightarrow g\) (almost uniformly) iff \(f=g(\text { a.e. }) ;\) similarly for convergence in measure.
    (c) If \(f\) and \(f_{n}\) are \(\mathcal{M}\) -measurable on \(A,\) then \(f_{n} \rightarrow f\) in measure on \(A\) iff
    (\forall \sigma>0) \quad \lim _{n \rightarrow \infty} m A\left(\rho^{\prime}\left(f, f_{n}\right) \geq \sigma\right)=0 .

    Exercise \(\PageIndex{12}\)

    Assuming that \(f_{n}: S \rightarrow\left(T, \rho^{\prime}\right)\) is \(m\) -measurable on \(A\) for \(n=1,2, \ldots,\) that \(m A<\infty,\) and that \(f_{n} \rightarrow f(a . e .)\) on \(A,\) prove the following.
    (i) Lebesgue's theorem: \(f_{n} \rightarrow f\) (in measure) on \(A\) (see Problem 11 ).
    (ii) Egorov's theorem: \(f_{n} \rightarrow f\) (almost uniformly) on \(A\).
    [Outline: (i) \(\left.f_{n} \text { and } f \text { are } \mathcal{M} \text { -measurable on } H=A-Q, m Q=0 \text { (Corollary } 1\right),\) with \(f_{n} \rightarrow f\) (pointwise) on \(H .\) For all \(i, k,\) set
    H_{i}(k)=\bigcap_{n=i}^{\infty} H\left(\rho^{\prime}\left(f_{n}, f\right)<\frac{1}{k}\right) \in \mathcal{M}
    by Problem 6 in \(\text { §1. Show that ( } \forall k) H_{i}(k) \nearrow H\); hence
    \lim _{i \rightarrow \infty} m H_{i}(k)=m H=m A<\infty ;
    (\forall \delta>0)(\forall k)\left(\exists i_{k}\right) \quad m\left(A-H_{i_{k}}(k)\right)<\frac{\delta}{2^{k}} ,
    proving \((\mathrm{i}),\) since
    \left(\forall n>i_{k}\right) \quad \rho^{\prime}\left(f_{n}, f\right)<\frac{1}{k} \text { on } H_{i_{k}}(k)=A-\left(A-H_{i_{k}}(k)\right) .
    (ii) Continuing, set \((\forall k) D_{k}=H_{i_{k}}(k)\) and
    D=A-\bigcap_{k=1}^{\infty} D_{k}=\bigcup_{k=1}^{\infty}\left(A-D_{k}\right) .
    Deduce that \(D \in \mathcal{M}\) and
    m D \leq \sum_{k=1}^{\infty} m\left(A-H_{i_{k}}(k)\right)<\sum_{k=1}^{\infty} \frac{\delta}{2^{k}}=\delta .
    Now, from the definition of the \(H_{i}(k),\) show that \(f_{n} \rightarrow f\) (uniformly) on \(A-D,\) proving (ii). \(]\)

    Exercise \(\PageIndex{13}\)

    Disprove the converse to Problem \(12(\mathrm{i})\).
    [Outline: Assume that \(A=[0,1) ;\) for all \(0 \leq k\) and all \(0 \leq i<2^{k},\) set
    g_{i k}(x)=\left\{\begin{array}{ll}{1} & {\text { if } \frac{i-1}{2^{k}} \leq x<\frac{i}{2^{k}}} \\ {0} & {\text { otherwise }}\end{array}\right.
    Put the \(g_{i k}\) in a single sequence by
    f_{2^{k}+i}=g_{i k} .
    Show that \(f_{n} \rightarrow 0\) in L measure on \(A,\) yet for no \(x \in A\) does \(f_{n}(x)\) converge as \(n \rightarrow \infty .]\)

    Exercise \(\PageIndex{14}\)

    Prove that if \(f: S \rightarrow\left(T, \rho^{\prime}\right)\) is \(m\) -measurable on \(A\) and \(g: T \rightarrow\left(U, \rho^{\prime \prime}\right)\) is relatively continuous on \(f[A],\) then \(g \circ f: S \rightarrow\left(U, \rho^{\prime \prime}\right)\) is \(m\)-measurable on \(A .\)
    [Hint: Use Corollary 4 in §1.]

    • Was this article helpful?