1. Limited time only! Sign up for a free 30min personal tutor trial with Chegg Tutors
    Dismiss Notice
Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Homework Help: Proving that a measurable function is integrable

  1. Jul 9, 2012 #1
    1. The problem statement, all variables and given/known data[/b]

    Let [itex]f:ℝ\rightarrowℝ[/itex] be measureable and [itex]A_{k}=\left\{x\inℝ:2^{k-1}<\left|f(x)\right|≤2^{k}\right\}[/itex], [itex]k\in[/itex] [itex]\mathbb{Z}[/itex].

    Show that [itex]f[/itex] is integrable only if [itex]\sum^{∞}_{k=-∞}2^{k}m(A_{k}) < ∞ [/itex].

    2. Relevant equations
    By the definition [itex]f[/itex] is integrable in ℝ if and only if [itex]f[/itex] is measurable and [itex]∫_{ℝ}\left|f\right|<∞[/itex].

    Now we know that [itex]f[/itex] is measureable, thus we should show that [itex]∫_{E}\left|f\right|≤ \sum^{∞}_{k=-∞}2^{k}m(A_{k}) < ∞ [/itex].

    3. The attempt at a solution

    Let [itex]f = \sum^{∞}_{k=-∞}f_{k}[/itex], we get [itex] \left|f \right| = \left| \sum^{∞}_{k=-∞} f_{k} \right| \Rightarrow \int_{ℝ}\left|f\right| = \int_{ℝ}\left|\sum^{∞}_{k=-∞}f_{k}\right| [/itex]

    (?)We also notice that [itex]\int_{ℝ}\left|\sum^{∞}_{k=-∞}f_{k}\right| = \sum^{∞}_{k=-∞}\int_{ℝ}\left|f_{k}\right| = 2 \sum^{∞}_{k=0}\int_{ℝ}\left|f_{k}\right| [/itex] (Beppo Levi's lemma?) (?)

    Let [itex]I_{k} = [2^{k-1}, 2^{k}][/itex]. The length of this interval is [itex]\ell(I_{k}) = 2^{k}-2^{k-1}=2^{k}(1-2^{-1}) = (1/2)2^{k}[/itex]

    We get [itex]m(A_{k})≤\ell(I_{k}) \Rightarrow \sum^{∞}_{k=-∞}m(A_{k})≤\sum^{∞}_{k=-∞}\ell(I_{k}) = \frac{1}{2}\sum^{∞}_{k=-∞}2^{k}[/itex].

    And thereby [itex]2\sum^{∞}_{k=-∞}m(A_{k}) = \sum^{∞}_{k=-∞}2m(A_{k}) ≤ \sum^{∞}_{k=-∞}2^{k}[/itex].

    Does my inference make any sense?
  2. jcsd
  3. Jul 9, 2012 #2


    User Avatar
    Science Advisor
    Homework Helper
    Gold Member

    What is [itex]f_k[/itex]? You didn't define it.

    Also, your assertion that


    is nonsense. [itex]A_k[/itex] is a subset of the domain of the function, whereas [itex]I_k[/itex] is a subset of the range. You certainly do not have [itex]A_k \subset I_k[/itex], which I suspect is the unstated assertion leading you to claim that [itex]m(A_{k})≤\ell(I_{k})[/itex]. E.g. if I put [itex]f(x) = 2^{k}[/itex] for all x, then [itex]A_k = \mathbb{R}[/itex] and thus [itex]m(A_k) = \infty[/itex].
    Last edited: Jul 9, 2012
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook