1
$\begingroup$

Let $(X,d)$ a metric space and $(x_n)_n\subset X$ a convergent sequence with limit $x=\lim_{n\to\infty}x_n$. Show that the set $\{x_n:n\ge1\}\cup\{x\}$ is compact.

Well, i'm with difficulties in start this question. I know that some set is compact (sequentially) if for any sequence in the set, there is a subsequence convergent in a point in the set. But i'm not understanding what the question wants and how to prove it.

Could someone help me? Even to start it?

$\endgroup$
5
  • 1
    $\begingroup$ Well, since you mention using sequences, any sequence in that space either contains $x$ infinitely often or finitely many times (possibly zero). Can you prove that there's a convergent subsequence in either case? $\endgroup$ Sep 15 '16 at 21:31
  • $\begingroup$ When I read $\{x_n:n\ge1\}\cup\{x\}$, the "$\cup$" had conspicuously less space to its left and right than what is standard. It turned out that that was because the $\{\text{curly braces}\}$ are outside of MathJax rather than inside it. I changed that. See my edit to the question. $\qquad$ $\endgroup$ Sep 15 '16 at 21:38
  • $\begingroup$ @TedShifrin Thankyou to the hint, friend, but i didn't understand the diference between x infinitely often and finitely many times... Can you discuss more about? $\endgroup$ Sep 15 '16 at 22:19
  • $\begingroup$ @MichaelHardy Oh, my friend. Thank You! I am newbie and desperate (LoL) But i understood the suggestion. Thank You. $\endgroup$ Sep 15 '16 at 22:21
  • 1
    $\begingroup$ @Lucas, if there are infinitely many entries of $x$ in your sequence, you have an obvious subsequence to choose. If there aren't, any subsequence of $\{x_n\}$ must converge (to a point of your set), so find one as a subsequence of your sequence. $\endgroup$ Sep 15 '16 at 22:25
2
$\begingroup$

Let $A=\{x\}\cup\{x_n:n\ge 1\}$; you’re supposed to prove that if the sequence $\langle x_n:n\ge 1\rangle$ converges to the point $x$, then the set $A$ is compact. I would use the definition of compactness: show that if $\mathscr{U}$ is an open cover of $A$, then $\mathscr{U}$ has a finite subset that still covers $A$.

HINT: Since $\mathscr{U}$ covers $A$, there is some $U_0\in\mathscr{U}$ such that $x\in U_0$. $U_0$ is open, so there is an $\epsilon>0$ such that $B_d(x,\epsilon)\subseteq U_0$. Now use the fact that $\langle x_n:n\ge 1\rangle$ converges to $x$ to show that all but finitely many points of $A$ are in $U_0$. It takes only finitely many more members of $\mathscr{U}$ to cover those finitely many points.

$\endgroup$
10
  • $\begingroup$ I don't think "hint" is the right word for what you wrote. It's not a hint; it's a sketch of the argument. $\qquad$ $\endgroup$ Sep 15 '16 at 21:39
  • 2
    $\begingroup$ @Michael: It is indeed a sketch of an argument. It is also a hint — an extended hint, but a hint nonetheless. For some students it’s pretty much an answer; for others it’s pretty much a minimal hint sufficient to do the job. For the most part I use the label HINT as a generic indication that the answer will require the reader to supply something. This usage is not peculiar to me: it can be found in textbook exercises, too. $\endgroup$ Sep 15 '16 at 21:44
  • $\begingroup$ Certainly I know one textbook (and I probably know others) where it is used in that way. But I think that word should be used for something much shorter but at the same time something with extensive implications. $\endgroup$ Sep 15 '16 at 22:29
  • $\begingroup$ @BrianM.Scott hello again (you already answered me in another question)! Thankyou for the support. I'm trying to unveil what you said and look what i've got: $\langle x_n:n\ge 1\rangle$ converges to $x$ means that exists an $\epsilon_2>0$ and an $N\in\Bbb{N}$ such that ${x_n}\subset{B_{\epsilon2}(x)}$ $\forall{n}$; $n>N$. Being $\epsilon_2\leq\epsilon$, implies that $B_{\epsilon2}(x)$, which contains infinite points of $\langle x_n:n\ge 1\rangle$, after $N$ points, is contained in ${B_{\epsilon}(x)}\subset{\mathcal{U_0}}\subset{\mathcal{U}}$. ... $\endgroup$ Sep 15 '16 at 23:22
  • $\begingroup$ If i'm in the right way, then i'm not understanding what i'm prooving here... $\endgroup$ Sep 15 '16 at 23:22

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.