0
$\begingroup$

I have the following proposition and proof, and I was just wondering if anyone could tell me if I was correct in my reasoning or not.

Let $f$ be a non negative integrable function. Show that the function $F$ defined by $F(x)=\int_{-\infty}^xf(t)dm(t)$ is continuous, where $m$ denotes the Lebesgue measure.

Proof: Since $f$ is integrable, we have that $\int_{-\infty}^{\infty}f(t)dm(t)=K$ for some $K \in \mathbb{R}$, and this gives us that $lim_{x\rightarrow \infty} \int_{-\infty}^xf(t)dm(t)=K$. From this, given $\epsilon > 0$, there exists an $N$ such that for all $x_1,x_2 \geq N$, we have that $K-\int_{-\infty}^{x_1}f(t)dm(t)<\frac{\epsilon}{2}$ and $K-\int_{-\infty}^{x_2}f(t)dm(t)<\frac{\epsilon}{2}$. So we have that $|F(x_1)-F(x_2)|=|\int_{-\infty}^{x_1}f(t)dm(t)-\int_{-\infty}^{x_2}f(t)dm(t)|$, and this is $\leq |K-\int_{-\infty}^{x_1}f(t)dm(t)|+$ $|K-\int_{-\infty}^{x_2}f(t)dm(t)|<\frac{\epsilon}{2}+\frac{\epsilon}{2}=\epsilon$.

I feel like all the components of the proof are there but it still feels a bit dodgy to me. What am I missing, and where are my errors in logic?

$\endgroup$
4
  • $\begingroup$ This is really wrong. You're not showing that the function is continuous at (say) $x=-1$. Also, I believe that in your question, it should be $F(x)$ not $F(t)$, since $t$ isn't an actual variable in your formula. $\endgroup$
    – Calvin Lin
    Jan 9, 2013 at 0:47
  • $\begingroup$ So how would you go about it? $\endgroup$ Jan 9, 2013 at 0:50
  • $\begingroup$ How would you show that $g(x)$ is a continuous function for all $ x \in (-\infty, \infty)$? Replace $g$ with $F$. $\endgroup$
    – Calvin Lin
    Jan 9, 2013 at 0:51
  • $\begingroup$ By picking a $\delta$ based on $\epsilon$ and $x$. I'm having problems finding a $\delta$ that works. $\endgroup$ Jan 9, 2013 at 0:56

1 Answer 1

3
$\begingroup$

Fix $x_0$. We have:

$$ |F(x) - F(x_0)| \le \int_{x_0}^x |f(t)| \, dm(t) $$

Since $f$ is integrable, $|f|dm$ is bounded and absolutely continuous with respect to $m$. Hence, for each $\varepsilon > 0$, we can find $\delta > 0$ so that:

$$ |x - x_0| < \delta \Rightarrow \int_{x_0}^x |f(t)| \, dm(t) < \varepsilon $$

$\endgroup$
2
  • $\begingroup$ How does $f$ integrable imply bounded? What about $f(x)=\frac{1}{\sqrt{x}}$? $\endgroup$ Jan 9, 2013 at 1:05
  • $\begingroup$ The measure defined by $|f|dm$ is bounded, not the function itself. $\endgroup$ Jan 9, 2013 at 1:06

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.