11
$\begingroup$

Let $f:(a,b)\to\mathbb{R}$ be continuous at all $x\in(a,b)$. If $\lim\limits_{x\to b^-}f(x)$ and $\lim\limits_{x\to a^+}f(x)$ exist in $\mathbb R$, how can we prove that $f$ is uniformly continuous on $(a,b)$?

This is my attempt, but I'm just not certain it is correct:

Let $\epsilon>0$. It is clear that $[a+\epsilon,b-\epsilon]\subseteq(a,b)$. It is known that $f(x)$ is continuous on $(a,b)$, so it is uniformly continuous on the bounded interval $[a+\epsilon,b-\epsilon]$. But, $[a+\epsilon,b-\epsilon] \iff (a,b)$. Thus $f$ is uniformly continuous on $(a,b)$.

Is my proof correct? Is there a better way?

$\endgroup$
5
  • 3
    $\begingroup$ As it stands, your proof is insufficient, and the equivalence $[a+\epsilon,b-\epsilon]\Leftrightarrow (a,b)$ makes no sense (and cannot be modified to make sense). $\endgroup$ Jan 2 '13 at 9:42
  • $\begingroup$ You haven't used the extra condition that $\lim_{x\to a^+} f(x)$ and $\lim_{x\to b^-} f(x)$ exist. Without them, the statement is false. (Take for example $f(x) = 1/(x-a)$.) $\endgroup$
    – mrf
    Jan 2 '13 at 9:45
  • 1
    $\begingroup$ I know what you're trying to do with the line $[a+\epsilon, b-\epsilon] \Leftrightarrow (a,b)$, but as Olivier said, it doesn't work. (Take $a = -\pi/2, b = \pi/2, f(x) = \tan(x)$ for a counterexample. Of course, the limits don't exist here as required.) Here's a hint: what do you know about continuous functions on closed intervals? Once you've answered that, can you find a large closed interval on which f is continuous? $\endgroup$
    – Billy
    Jan 2 '13 at 9:47
  • 1
    $\begingroup$ You have made mention of the theorem according to which a continuous function on a closed bounded interval $[c,d]$ is automatically uniformly continuous. Instead of trying to use this property on subintervals of $(a,b)$, use the hypothesis on $f$ to extend it continuously to some larger interval, and use the theorem. $\endgroup$ Jan 2 '13 at 9:48
  • $\begingroup$ @Olivier Ahhh... A neat alternative to the tedious epsilon-delta approach. $\endgroup$
    – Did
    Jan 2 '13 at 9:53
8
$\begingroup$

Your proof is incorrect, because $f$ being uniformly continuous on all closed subintervals of $(a,b)$ doesn't imply it is so on $(a,b)$ itself.

Here is one way to do it. Extend the function $f$ to the function $g:[a,b]\to\mathbb R$ with $g(a)=\lim_{x\to a^+} f(x)$ and $g(b)=\lim_{x\to b^-}f(x)$. Then $g$ is continuous on $[a,b]$. It follows that $g$ is uniformly continuous on $[a,b]$ so that $f$ is uniformly continuous on $(a,b)$.

$\endgroup$
2
  • $\begingroup$ Hi, thanks for the reply! $\endgroup$
    – Harold
    Jan 2 '13 at 11:19
  • $\begingroup$ @Harold very sorry to re-open it, but if you can, why from "g is uniformly continuous on [a,b] so that f is uniformly continuous on (a,b)."? $\endgroup$ Feb 23 '15 at 15:37

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.