Mathematics Stack Exchange is a question and answer site for people studying math at any level and professionals in related fields. Join them; it only takes a minute:

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

could anyone help to show that $[0,1]^{\mathbb{N}}$ with respect to the box topology is not compact? Thank you!

share|cite|improve this question
up vote 11 down vote accepted

There’s absolutely nothing wrong with showing non-compactness of $X = \square_{k=0}^\infty [0,1]$ directly by looking at open covers, but there are other ways as well. For instance:

If $X = {\Large \square}_{k=0}^\infty [0,1]$ were compact, its closed subspace $ {\Large \square}_{k=0}^\infty \{0,1\}$ would be compact, but it’s not hard to show that $ {\Large \square}_{k=0}^\infty \{0,1\}$ is an infinite, closed, discrete set in $X$ and therefore cannot be compact.

Even simpler:

For $n\in\mathbb{N}$ let $x_n \in X$ be the point such that $x_n(n) = 1$ and $x_n(k) = 0$ if $k\ne n$. Now consider the set $A = \{x_n:n\in\mathbb{N}\}$. It’s infinite, so if $X$ were compact, $A$ would have a limit point in $X$. But it’s not hard to show that $A$ is a closed, discrete subset of $X$ and therefore has no limit point in $X$.

share|cite|improve this answer
Just so nobody else gets confused like I did, the square box means "Cartesian product with the box topology", not some kind of missing or mis-rendered character in your browser. – Ted Aug 28 '11 at 6:42
@Ted: Sorry about that; I used to work a bit with box products and got used to that notation. – Brian M. Scott Aug 28 '11 at 8:36
@BrianM.Scott How to prove that $ {\Large \square}_{k=0}^\infty \{0,1\}$ is closed in $ {\Large \square}_{k=0}^\infty [0,1]$? – Twink May 16 '15 at 3:59
@user: For $n,k\in\Bbb N$ let $U_n(k)=(0,1)$ if $n=k$ and $[0,1]$ otherwise. For $n\in\Bbb N$ let $U_n=\prod_{k\in\Bbb N}U_n(k)$. Then $U_n$ is open in ${\Large\square}_{k\in\Bbb N}[0,1]$, and $$\left({\Large\square}_{k\in\Bbb N}[0,1]\right)\setminus{\Large\square}_{k\in\Bbb N}\{0,1\}=\bigcup_{n\in\Bbb N}U_n\;.$$ – Brian M. Scott May 16 '15 at 4:26
@user: $[0,1]^{[0,1]}$ with the box topology? Yes. Or you could you observe that $[0,1]^{[0,1]}$ has a closed subspace homeomorphic to ${\Large\square}_{k\in\Bbb N}[0,1]$, and since the latter is not compact, the former cannot be either. – Brian M. Scott Jun 2 '15 at 15:52

Let $a=(a_1,a_2,\dots)$ be a sequence consisting of only $0$'s and $1$'s and define $U_{an}=[0,\frac{3}{4})$ if $a_n=0$ and $U_{an}=(\frac{1}{4},1]$ if $a_n=1$. Also, define $U_a=\prod_{n=1}^{\infty} U_{an}$.

Exercise 1: Prove that $U_a$ is an open subset of $[0,1]^{\mathbb{N}}$ in the box topology but that it is not an open subset of $[0,1]^{\mathbb{N}}$ in the product topology.

Exercise 2: Prove that the collection of subsets of $[0,1]^{\mathbb{N}}$ of the form $U_a$ where $a=(a_1,a_2,\dots)$ is a sequence consisting of only $0$'s and $1$'s is an open cover of $[0,1]^{\mathbb{N}}$ in the box topology.

Exercise 3: Does this open cover have a finite subcover?

The following problems should be of additional interest:

Problem 1: Let $\{X_n\}_{n\in\mathbb{N}}$ be a collection of topological spaces. If the box topology on $\prod_{n=1}^{\infty} X_n$ is a compact topological space, what can you deduce about the $X_n$'s? In other words, how do the topologies on the individual $X_n$'s affect the compactness of $\prod_{n=1}^{\infty} X_n$ in the box topology?

Problem 2: In the context of Problem 1, if $\prod_{n=1}^{\infty} X_n$ is compact in the box topology, then prove that $X_n$ is compact for all $n\in\mathbb{N}$.

Problem 3: A topological space $X$ is said to be countably compact if every open cover of $X$ has a countable subcover. Is $[0,1]^{\mathbb{N}}$ countably compact in the box topology?

I hope this helps!

share|cite|improve this answer
Problem 1 is best set in the context of $T_1$-spaces; otherwise a complete answer gets a little messy. Assuming that one already has the Tikhonov theorem, Problem 2 has an easy solution that doesn’t require knowing the result of Problem 1. (Not a complaint, just an observation.) – Brian M. Scott Aug 28 '11 at 8:57
@Brian Of course, you are right. I did not suggest that Problem 2 requires Problem 1; I only suggested that the notation of Problem 2 is borrowed from Problem 1 (thus, "In the context of Problem 1"). Also, I do not think one needs Tikhonov's theorem to solve Problem 2; one can note that the projection map $\prod_{n=1}^{\infty} X_n\to X_i$ is continuous for all $i\in\mathbb{N}$. Problem 1 is more of an open-ended problem; the OP is free to interpret the problem whichever way he/she chooses. If he/she chooses to assume the $T_1$-axiom, then this is fine. – Amitesh Datta Aug 28 '11 at 23:09

$$\left\{\displaystyle\prod_{n\in \mathbb{N}} I_n : I\in \left\{[0,\frac23),(\frac13,1]\right\}^\mathbb{N}\right\}$$

is an open cover of $[0,1]^{\mathbb{N}}$ with the box topology that does not have a finite subcover.

share|cite|improve this answer

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

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