Good morning, I'm doing Problem III.2.12 from textbook Analysis I by Amann.

enter image description here Here $\overline A$ is the set of all accumulation points of $A$ or equivalently the closure of $A$.

Could you please verify if my proof look fine or contains logical gaps/errors? Any suggestion is greatly appreciated.

My attempt:

$\Longrightarrow$ (Indirect proof): Assume the contrary that $f(\overline A) \not \subseteq \overline{f(A)}$. Then there is $y \in f(\overline A)$ such that $y \notin \overline{f(A)}$. Because $y \in f(\overline A)$, there is $x \in \overline A$ such that $f(x) =y$. Because $y \notin \overline{f(A)}$, there is a neighborhood $U$ of $y$ such that $U \cap f(A) = \emptyset$. Because $f$ is continuous, there is a neighborhood $U'$ of $x$ such that $f(U') \subseteq U$. As such, $f(U') \cap f(A) = \emptyset$ and thus $U' \cap A = \emptyset$, which contradicts $x \in \overline A$.

$\Longrightarrow$ (Direct proof): For $y \in f(\overline A)$, there is $x \in \overline A$ such that $f(x) = y$. Because $f$ is continuous, if $U$ is a neighborhood of $y$ then there is a neighborhood $U'$ of $x$ such that $f(U') \subseteq U$. It follows from $x \in \overline A$ that $U' \cap A \neq \emptyset$, so $f(U') \cap f(A) \neq \emptyset$. As such, $U \cap f(A) \neq \emptyset$ and hence $y \in \overline{f(A)}$.

$\Longleftarrow$: Let $A \subseteq Y$ be closed in $Y$ and $B = f^{-1}(A) \subseteq X$. Assume the contrary that $B$ is not closed in $X$, then $C = \overline B - B \neq \emptyset$ and $C \cap B = \emptyset$. We have $f(C \cup B) = f(\overline B) \subseteq \overline {f(B)} = \overline A = A$, so $(C \cup B) \subseteq f^{-1}(A) = B$, which contradicts $C \neq \emptyset$ and $C \cap B = \emptyset$. As such, $B$ is closed and thus $f$ is continuous.

  • 1
    $\begingroup$ Your reasoning looks sound to me! $\endgroup$ – Math1000 Sep 7 at 6:41
  • $\begingroup$ Thank you so much for your help @Math1000 :)))))) $\endgroup$ – Abstract Analysis Sep 7 at 6:45

$f[B]=f[f^{-1}[A]] \subseteq A$, not $=A$ (unless $f$ is surjective), but that's a minor thing; you only need the inclusion anyway.

The first part can be done easier: As $\overline{f[A]}$ is closed, $f^{-1}[\overline{f[A]}]$ is closed by continuity and contains $A$, i.e. $$A \subseteq f^{-1}[f[A]] \subseteq f^{-1}[\overline{f[A]}]$$

so $$f[\overline{A}] \subseteq \overline{f[A]}$$ is immediate. This does require you know that $f$ continuous means that the inverse images of closed sets are closed. (As for open sets.)

  • $\begingroup$ Please check if my understanding is correct! It follows from $A \subseteq f^{-1}[f[A]] \subseteq f^{-1}[\overline{f[A]}]$ that $\overline{A} \subseteq \overline{f^{-1}[\overline{f[A]}]}$. Because $f^{-1}[\overline{f[A]}]$ is closed, $\overline{f^{-1}[\overline{f[A]}]} = f^{-1}[\overline{f[A]}]$. As such, $\overline A \subseteq f^{-1}[\overline{f[A]}]$ and thus $f[\overline A] \subseteq \overline{f[A]}$. $\endgroup$ – Abstract Analysis Sep 7 at 8:31
  • $\begingroup$ To remove ambiguity, I prefer $f[A]$ than $f(A)$ for the induced set-value function from $f$ as you did. But not many people use this notation and it makes the notation complicated in some cases. Your approach is much more elegant beautiful than mine. $\endgroup$ – Abstract Analysis Sep 7 at 8:34
  • $\begingroup$ Thank you so much for correcting my critical logical mistake! $\endgroup$ – Abstract Analysis Sep 7 at 8:38
  • $\begingroup$ @LeAnhDung yes, that’s right. I assumed the intermediate step. $\endgroup$ – Henno Brandsma Sep 7 at 8:47

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.