MathOverflow is a question and answer site for professional mathematicians. 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

Let $X$ be a topological space with the following property: for any open subset $A$ of $X$ there is a collection of clopen subsets $\{A_{\alpha} : \alpha\in S\}$ such that $\overline{A}=\overline{\bigcup A_{\alpha}}$ (where $\overline{A}$ denotes the closure of $A$). Can anyone find an example of a space which has the above property but is not zero-dimensional?

share|cite|improve this question
Motivation? This seems like a homework problem . . . – Noah Schweber Nov 23 '12 at 21:02
You haven't specified which definition of dimension you're using. Have you looked in the standard sources for whichever definition you're using? I agree with Noah, the question seems poorly motivated. – Ryan Budney Nov 23 '12 at 21:47
Clarification... One definition for "zero-dimensional" (small inductive dimension, I guess) says: there is a base for the topology consisting of clopen sets. We can rephrase that as: for any open set $A$ there is a colletion of clopen sets $\{A_{\alpha}: \alpha\in S\}$ such that ${A}={\bigcup A_{\alpha}}$. So clearlly such a space has the property in the question. But maybe adding the closure gets us additional spaces. – Gerald Edgar Nov 23 '12 at 21:56
up vote 4 down vote accepted

Any space which contains a dense set of isolated points will have the property. It is easy to construct (e.g. as subspaces of $\mathbb{R}^n$) such spaces with arbitrarily large dimension. For a one dimensional explicit construction see B.M.Scott's answer to this question:

share|cite|improve this answer
Dear Ramiro What you think about the following question? Is there a completely regular Hausdorff space $X$ which has my property(in main questio) but is not zero-dimensional? – Ali Nov 24 '12 at 12:55
Ali, subspaces of the plane are completely regular (in fact normal) Hausdorff spaces. – Ramiro de la Vega Nov 24 '12 at 13:08
Any space which contains a dense set of isolated points has my property but may be a zero-dimensional space so is not my answer (e.g, $\beta N$ ). – Ali Nov 24 '12 at 20:38
Note that Zero-dimesional is a $T_{1}$ space which has a base contains clopen subsets. – Ali Nov 24 '12 at 20:39
Ali, You must have misunderstood the answer that Ramiro links to. It gives an example of a subset in $\mathbb{R}^2$ that contains an interval and also a dense set of isolated points. (Take a line segment $I$ in $\mathbb{R}^2$, and add a sequence of points in $\mathbb{R}^2\setminus I$ that accumulates everywhere on $I$.) – Lasse Rempe-Gillen Nov 25 '12 at 16:31

Yes, there are spaces where this property holds which are not zero-dimensional: consider the two-point space $\lbrace a, b\rbrace$ with topology $\lbrace \lbrace a\rbrace, \lbrace a, b\rbrace, \emptyset \rbrace$. Then this space clearly does not have a basis of clopen sets, so is not zero-dimensional, but the only closed sets are $\emptyset$ and $\lbrace a, b\rbrace$ which are both the closures of clopen sets.

Note that a strengthening of your property - where we require the clopen $A_\alpha$ to be subsets of $A$ - does not hold in this case, and I suspect that stronger version is equivalent to being zero-dimensional. [EDIT: as Ramiro points out, this is not the case; see comments]

Also, note that there are multiple definitions of "zero-dimensional," so you should specify which one you mean.

(I am still curious what the motivation for this problem is.)

share|cite|improve this answer
Spaces with a dense set of isolated points also have the stronger property you propose. And there are plenty of non-zero-dimensional spaces with a dense set of isolated points. – Ramiro de la Vega Nov 24 '12 at 3:50
Dear Friends Thank you very much of your answers. A $T_{1}$-space $X$ which has a base consisting of clopen sets is called zero-dimensional space. – Ali Nov 24 '12 at 4:39
Given that you want your space to be $T_1$, my example doesn't apply; Ramiro's, on the other hand, appears to exactly answer your question. – Noah Schweber Nov 24 '12 at 6:08

It is Easy to see that $X$ has my property if and only if $\beta X$ has my property. So if we consider Dowkers example $Y$ then $Y$ is a zero-dimensional where $\beta Y$ is not zero-dimensional (see general topology book, Ryszard Engelking). On the other hand $\beta Y$ has my property.

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.