Definitions. Suppose $X$ is a topological space.

  • $w(X)=\min\{|\mathcal B|:\mathcal B$ is a base for $X\}+\omega$
  • $e(X)=\sup\{|D|:D\subseteq X$ is closed and discrete$\}+\omega$
  • $K(X)$ is the collection of all compact subsets of $X$
  • $\mathbb R ^*=\beta\mathbb R \setminus \mathbb R$

The collection of all open subsets of $\mathbb R$ has cardinality $\mathfrak c$. The canonical basis for $\beta \mathbb R$ consists of sets of the form

$$\{p\in\beta\mathbb R :(\exists L\in p)(L\subseteq U)\}$$ where $U$ is open in $\mathbb R$. Thus $w(\beta\mathbb R) \leq \mathfrak c$ ( I suspect equal).

Now I appeal to a theorem, which may or may not be a deep result (I haven't read the proof) in cardinal invariants.

Theorem. If $X$ is compact $T_1$ then $|K(X)|\leq 2^{e(X)\cdot w(X)}$.

Note that $e(\mathbb R ^*)=\omega$ (really finite, but in the definition we require that it be at least $\omega$): If $D$ is an infinite discrete closed subspace of $\mathbb R^*$, then $D$ contains a copy of $\beta\omega$, contradicting $D$ discrete.

Combining everything we have $|K(\mathbb R ^*)|\leq 2^{\omega\cdot \mathfrak c}=2^\mathfrak c$ (whereas $|\mathcal P (\mathbb R^*)|=2^{2^\mathfrak c}$).


1) Is my calculation correct?

2) Is there an easier way to get my final conclusion? I feel there must be, but maybe not. We know $\beta\mathbb R$ has a countable dense subset, but I didn't see how to use that.

Edit: I might be an idiot. Why does the following not work: $\beta\mathbb R$ has a countable dense subset $\mathbb Q$, and we can uniquely identify an open subset of $\beta \mathbb R$ by its intersection with $\mathbb Q$. Thus there are only $2^\omega$ open subsets of $\beta\mathbb R$, whence there are also only $2^\omega$ closed subsets of $\beta\mathbb R$... clearly false.


Is there an easier way to get my final conclusion?

Let $o(X)$ denote the cardinality of the topology of the space $X$ (i.e., it is the number of open subsets of $X$.) The same notation is used in Juhasz' book Cardinal Functions in Topology - Ten Years Later.

Then we have inequality $o(X)\le 2^{w(X)}$, which follows from the fact that every open set can be obtained as a union of some system of basic sets.

If we use $w(\beta\mathbb R)\le\mathfrak c$, then we get $$o(\beta\mathbb R)\le 2^{\mathfrak c}.$$

So there are at most $2^{\mathfrak c}$ open subsets in $\beta\mathbb R$. Clearly, the number of closed subsets is the same. And since we are working with subsets of a compact space, compact subsets are precisely closed subsets, so we get $|K(\beta\mathbb R)|\le 2^{\mathfrak c}$.

Since weight of a subspace is less or equal to the weight of the whole space, we also have $w(\mathbb R^*)\le w(\beta\mathbb R) \le \mathfrak c$ and we can use the same argument to get $$|K(\mathbb R^*)|=o(\mathbb R^*) \le 2^{\mathfrak c}.$$

On the other hand, we have $o(X)\ge |X|$ for any $T_1$ space. Since $|\beta\mathbb R|=2^{\mathfrak c}$ we get $$o(\beta\mathbb R)=2^{\mathfrak c}.$$ For the proof of cardinality of $\beta\mathbb R$ see: Stone–Čech compactification of $\mathbb{N}, \mathbb{Q}$ and $\mathbb{R}$

  • $\begingroup$ Thanks, I thought of that before sleeping last night, but I will give credit for the answer anyway! $\endgroup$ – Forever Mozart Jun 30 '14 at 14:50

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.