2
$\begingroup$

Consider the metric space $\mathbb{R}$ equipped with the standard distance metric. Let $S$ be a set of rational numbers in the open interval $(a,b)$ where $a$ and $b$ are irrational. Prove that $S$ is closed in the set of rational numbers $\mathbb{Q}$.
Attempt:

$S^c = \{x \in \mathbb{Q} : x \leq a \} \cup \{x \in \mathbb{Q} : x \geq b \}$

Let $x < a$. Choose $\epsilon = a-x$. We can find a $N$ such that $1/N < \epsilon$.
This implies that $x + 1/N$ is rational and less than $a$. A similar argument can be made for $x > b$.
So we see, that $\forall x \in S^c, \exists \epsilon > 0$ such that $B_{\epsilon}(x) \subset S^c$. So $S^c$ is open and so $S$ is closed.

Does this make sense?

$\endgroup$
  • $\begingroup$ This is correct (and presumably the intended answer). $\endgroup$ – Mees de Vries May 2 '16 at 11:20
  • $\begingroup$ "Closed" in which ambient space X and for which topology on X? $\endgroup$ – Did May 2 '16 at 11:39
  • $\begingroup$ @MeesdeVries: It's close, but not quite there. $\endgroup$ – Cameron Buie May 2 '16 at 12:37
  • 1
    $\begingroup$ Please answer the question that @Did asked and include it in your question. $\endgroup$ – zhw. May 2 '16 at 16:57
  • 1
    $\begingroup$ Except that now, "Consider the metric space $\mathbb{R}$ equipped with the standard distance metric" and "Prove that $S$ is closed in the set of rational numbers $\mathbb{Q}$" are contradictory. If the ambient space is $X$ and if one is given $S\subset Y\subset X$, please explain what you mean by "$S$ is closed in $Y$". $\endgroup$ – Did May 2 '16 at 17:16
1
$\begingroup$

One thing that I'd adjust is this: you never used the fact that $a,b$ are irrational. In particular, that lets you conclude that $a,b\notin S^c,$ so that $$S^c=\{x\in\Bbb Q:x<a\}\cup\{x\in\Bbb Q:x>b\}.$$ (Do you see why this is important?)

Another thing I'd adjust is the part with $N$--it doesn't really help you show that $B_\epsilon(x)\subseteq S^c.$ Rather, take $\epsilon=a-x$ as you did, and take any $y\in B_\epsilon(x),$ meaning that $y\in\Bbb Q$ and $|y-x|<\epsilon.$ In particular, since $y-x\le|y-x|,$ it then follows that $y<a.$ (Do you see how?) Consequently, $y\in S^c,$ and since $y\in B_\epsilon(x)$ was arbitrary, then $B_\epsilon(x)\subseteq S^c,$ as desired.

$\endgroup$
  • $\begingroup$ That's definitely a tighter proof. Thanks. $\endgroup$ – sid May 2 '16 at 13:31
  • $\begingroup$ @Cameron Buie can you please provide a hind about compactness of this set? $\endgroup$ – gaurav saini Sep 23 '19 at 12:23
  • $\begingroup$ @gaurav: What do you mean? This question doesn't have anything to do with compactness. $\endgroup$ – Cameron Buie Sep 23 '19 at 12:52
0
$\begingroup$

Here is another proof of this.

Theorem. Let $S$ be a subset of the metric space $M$. $X \subseteq S$ is closed in $S$ if and only if there is an open set $A$ in $M$ such that $X=S\cap A$.

Using this theorem for your example, $A=[a,b]$ is closed in $\mathbb{R}$. Also, $X=\{y\,|\,y\in\mathbb{Q}\,\text{and}\,y\in(a,b)\}$. We see that $X=\mathbb{Q}\cap A$ because $a$ and $b$ are irrational numbers. So $X$ is closed in $\mathbb{Q}$.

$\endgroup$

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.