3

Let $A$ be a rectangle in $\mathbb{R^k}$; let $B$ be a rectangle in $\mathbb{R^n}$; let $Q=A\times B$. Let $f: Q\to \mathbb{R}$ be a bounded function. Show that if $\int_Q f$ exists, then

$$\int_{y\in B}f(x,y)$$ exists for $x\in A-D$, where $D$ is a set of measure zero in $\mathbb{R^k}$.

I'm having difficulty proving this.

I think I might have to use these theorems in the proof.

enter image description here

enter image description here

My work:

Let $D$ be a set of measure zero in $\mathbb{R^k}$, and $x\in A-D$.

Since $f$ is integrable on $Q$, $f$ is continuous except in a measure zero set, say $E$.

Claim: Let $B$ be a rectangle in $\mathbb{R^n}$. If $D$ is a set of measure zero in $\mathbb{R^k}$, then $D\times B$ is a set of measure zero in $\mathbb{R^{k+n}}$.

Proof of Claim: Let $\epsilon\gt 0$ be given.Since $D$ is a set of measure zero in $\mathbb{R^k}$, there exists a countable set of rectangles $\{Q_i\}$ whose union covers $D$ and the sum of the volumes is less than $\epsilon/v(B)$, where $v(B)$ is the volume of $B$. Now consider the countable set of rectangles $Q_i \times B$. Then the union of all of these rectangles cover $D\times B$, and $\sum v(Q_i\times B)=\sum v(Q_i)\times v(B)\lt \epsilon$. QED.

Hence by the claim above, $E \cup (D\times B)$ is a set of measure zero in $\mathbb{R^{k+n}}$. Hence, $f$ is continuous almost everywhere when $x\in A-D$.

Now for a fixed $x\in A-D$, $f$ is continuous almost everywhere in $\mathbb{R^{k+n}}$. However, to complete the proof of this problem, I need to show that $f(x,y)$, for a fixed $x\in A-D$ is continuous almost everywhere in $\mathbb{R^n}$, but I don't know how to show this part.

I would greatly appreciate any hints, suggestions or solutions.

  • This looks like Fubini's theorem (for Lebesgue integrals) to me, did you have a look at that? Your approach is wrong because you assume the existence of the set D, but you have to show that such a set exists. – Martin R Oct 19 '15 at 06:43
  • @MartinR I added the Fubini's Theorem as given in the text, but I'm not sure how this can help prove this. And why do I have to show that such a set exists? Isn't the problem saying that if $D$ is any set of measure zero in $R^k$ then $f(x,y)$ for $x\in A-D$ would be integrable over $B$? – nomadicmathematician Oct 19 '15 at 12:27
  • In which context is this? In the context of measure theory, an integrable function can be nowhere continuous. In $\mathbb{R}$, the function $f(x)=1$ if $x\in \mathbb{Q}$ and $f(x)=0$ if $x\in \mathbb{R}/\mathbb{Q}$ is measurable, and integrable (the integral is zero), and this function is nowhere continuous. – Nate River Oct 21 '15 at 11:29
  • @NateRiver I put up the context in the question, I think it's Riemann. – nomadicmathematician Oct 21 '15 at 11:46

0 Answers0