3
$\begingroup$

Let $A$ be a compact set contained in a open set $\Omega$. Then there exists $\epsilon > 0$ such that $A \subset \cup_{x \in A} B(x,\epsilon) \subset \Omega$.

How can I prove it? I have tried it by proving compactness implies total boundedness. But that doesn't work properly.Please help me.

Thank you in advance.

$\endgroup$

2 Answers 2

3
$\begingroup$

Assume this isn't true. Then for every $n\in \mathbb{N} $ there exists some $x_n\in A$ such that $B (x_n,\frac{1}{n})\cap \Omega^c\neq \emptyset $. Consider the sequence $x_n $. By compactness of $A $, there exists a convergent subsequence of $x_n $. Call the limit $x $. Then $x\in A $, but $x $ is a boundary point of $\Omega $ and hence not an element of $\Omega $, so we have our contradiction.

$\endgroup$
3
  • $\begingroup$ Why does $x \in \partial A$? $\endgroup$ Commented Oct 20, 2017 at 13:12
  • $\begingroup$ @ArnabChatterjee. What is important is that it is in the boundary of $\Omega $. To see that that is true, note that every open ball containing $x$ also contains infinitely many $x_n $ by definition. Then taking open balls of radius small enough gives the result. $\endgroup$ Commented Oct 20, 2017 at 14:05
  • $\begingroup$ What I thought is that if $x \in A$ then clearly $x \in \Omega$. Then there exists $\delta>0$ such that $B(x;\delta) \subset \Omega$. Now if $\{x_{k_n} \}$ be the subsequence converging to $x$ then there exists $p \in \mathbb N$ such that for all $n \geq p$, $d(x_{k_n},x) < \delta$. Take $q \geq p$ such that $\frac {1} {k_q} < \delta$. Then if I take any $y \in B({x_{k_q}}, {\frac {1} {k_q}}) \cap \Omega^{c}$ then by triangle inequality $d(x,y) < 2\delta$ and it is true for any $\delta' < \delta$. Hence $x=y$. But then $x \in \Omega^c$!! $\endgroup$ Commented Oct 20, 2017 at 15:36
3
$\begingroup$

Let $f(x)= d(x,\partial \Omega)$ defined on $A$. Since $A$ is compact and $f$ is continuous on $A$, $f$ has a minimum value $\epsilon_0\geq 0$. If $\epsilon_0 = 0$, there exists $x_0\in A$ such that $x_0 \in \partial \Omega$, which cannot be true by hypothesis.

Now, we set $\epsilon = \frac{\epsilon_0}{2}$, and we have $A \subset \cup_{x \in A} B(x,\epsilon) \subset \Omega$.

$\endgroup$

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .