Hi¡ I have a problem with the next exercise.

Let $(X,\tau)$ be a topological space. Prove that $\mathcal{U}\subset X$ is open in $X$ if and only if for all $\mathcal{A}\subset X$ holds that $\text{cl}(\mathcal{U}\cap\text{cl}(\mathcal{A}))=\text{cl}(\mathcal{U}\cap \mathcal{A})$ where $\text{cl}()$ is the closure of a set and $\text{der}()$ is the derived set.

My attempt:


Let $\mathcal{A}\subset X$


Then, $\mathcal{U}\cap\mathcal{A}\subseteq\mathcal{U}\cap\text{cl}(\mathcal{A})$ and therefore, $\text{cl}(\mathcal{U}\cap\mathcal{A})\subseteq\text{cl}(\mathcal{U}\cap\text{cl}(\mathcal{A}))$

For the another contention, the same:


Let $x\in\text{cl}(\mathcal{U}\cap\text{cl}(\mathcal{A}))$, then, $x\in\mathcal{U}\cap\text{cl}(\mathcal{A)}$ or $x\in\text{der}(\mathcal{U}\cap\text{cl}(\mathcal{A}))$

If $x\in\mathcal{U}\cap\text{cl}(\mathcal{A})$, then, $x\in \mathcal{U}$ and ($x\in \mathcal{A}$ or $x\in\text{der}{\mathcal{A}}$)

If $x\in \mathcal{A}$ and $x\in \mathcal{U}$, then, clearly, $x\in\text{cl}(\mathcal{U}\cap\mathcal{A})$

If $x\in\mathcal{U}$ and $x\in\text{der}(\mathcal{A})$, then, $x\in\text{cl}(\mathcal{U})\cap\text{cl}(\mathcal{A})$, but, here, again, I'm very stuck.

For the $\Leftarrow)$ I don´t know how can I do it. I need a hint, or more.

Thanks in advance.


I think it's best to use the characterisation

$x \in \overline{A}$ iff for all open sets $O$ that contain $x$, $O \cap A \neq \emptyset$, instead of the formula $\overline{A} = A \cup A'$ (where the latter is the derived set of $A$, its limit points), and doing a case analysis.

I wrote up a complete proof here, using this characterisation.

Inspired by the erroneous proof using sequences by inquisitive, I came up with a proof using nets; I will not give it (except upon request), because most courses do not treat nets at all.


continuing your proof you can say that,

since $x \in \overline{\mathcal{A}}$, $\exists$ a sequence $\{x_i\} \in \mathcal{A},\ \{x_i\} \rightarrow x$,

and since $x \in \mathcal{U}$ and $\mathcal{U}$ is open we can find a $N \in \mathbb{N}$ st $\forall n > N, x_n \in \mathcal{U}$, thus we have a sequence $\{x_i , i>n \} \in \mathcal{A} \cap \mathcal{U} $ converging to $x$.

thus, $x \in \overline{\mathcal{A} \cap \mathcal{U}} $

for the reverse you can prove if $\overline{\mathcal{A} \cap \mathcal{U}} = \mathcal{A} \cap \overline{\mathcal{U}} \ \ \forall \ \mathcal{A} \subset X$ then, $ X \setminus \mathcal{U}$ is closed.

to prove this, let $y \in der(X \setminus \mathcal{U})$ s.t $y\notin X \setminus \mathcal{U}$

then $\exists$ a sequence $\{y_i\} \in X \setminus \mathcal{U}$ converging to $y$

let $\mathcal{A} = \{y_i , i\in \mathbb{N}\}$

now we know that $y\in \overline{\mathcal{A}}$

Since $\mathcal{A} \subset X \setminus \mathcal{U}$, therefore, $\mathcal{U}\cap\mathcal{A} = \overline{\mathcal{U}\cap\mathcal{A}} = \varnothing$

But, $y \in \mathcal{U} \cap \overline{\mathcal{A}}$, which is a contradiction.

Hence, $y \in X \setminus \mathcal{U} \implies \mathcal{U}$ is open.

  • $\begingroup$ There does not need to be such a sequence ,so this is not a general proof. $\endgroup$ – Henno Brandsma Mar 2 '17 at 19:54

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.