5
$\begingroup$

More homework help. Given the function $f:A \to B$. Let $C$ be a subset of $A$ and let $D$ be a subset of $B$.

Prove that:

$C$ is a subset of $f^{-1}[f(C)]$

So I have to show that every element of $C$ is in the set $f^{-1}[f(C)]$

I know that $f(C)$ is the image of $C$ in $B$ and that $f^{-1}[f(C)]$ is the pre-image of $f(C)$ into $A$. Where I'm stuck is how to use all of this information to show/prove that $C$ is indeed a subset.

Do I start with an arbitrary element (hey, let's call it $x$) of $C$? and then show that $f^{-1}[f(x)]$ is $x$? I could use a little direction here... Thanks.

$\endgroup$
12
$\begingroup$

Since you want to show that $C\subseteq f^{-1}\big[f[C]\big]$, yes, you should start with an arbitrary $x\in C$ and try to show that $x\in f^{-1}\big[f[C]\big]$. You cannot reasonably hope to show that $f^{-1}\big[f[\{x\}]\big]=x$, however: there’s no reason to think that $f$ is $1$-$1$, so there may be many points in $A$ that $f$ sends to the place it sends $x$.

Let $x\in C$ be arbitrary. For convenience let $E=f[C]\subseteq B$. Now what elements of $A$ belong to the set $f^{-1}\big[f[C]\big]=f^{-1}[E]$? By definition $f^{-1}[E]=\{a\in A:f(a)\in E\}$. Is it true that $f(x)\in E$? If so, $x\in f^{-1}[E]=f^{-1}\big[f[C]\big]$, and you’ll have shown that $C\subseteq f^{-1}\big[f[C]\big]$.

$\endgroup$
  • $\begingroup$ So it was all about using the definition of the preimage along the way to prove my assumption. It is always so clear in hindsight. Hopefully this gets easier before I get to Abstract Algebra and Real Analysis... $\endgroup$ – Ben Anderson Apr 1 '13 at 5:11
  • 1
    $\begingroup$ @Ben: You’ll find that a very large fraction of the more elementary exercises really do just boil down to working through the relevant definitions. You generally have to get a bit deeper into a subject before you encounter problems that require much more than that. $\endgroup$ – Brian M. Scott Apr 1 '13 at 5:13
  • $\begingroup$ @BrianM.Scott: I have added my answer to a question recently marked as a duplicate of this one. If you feel it is too similar to yours, I will remove it. $\endgroup$ – robjohn Sep 1 '15 at 21:49
  • $\begingroup$ @robjohn: Both the difference in style and the inclusion of the example make it worth keeping. (Besides, this made me catch a bad typo in my answer!) $\endgroup$ – Brian M. Scott Sep 2 '15 at 4:26
3
$\begingroup$

Copied from my answer to I am issues with proving the following problem: $f^{-1}(f(A)) ⊃ A$, which was closed as a duplicate of this question just before I posted it.


$$ f^{-1}\left(f(A)\right)=\left\{x:f(x)\in f(A)\right\}\tag{1} $$ Note that if $x\in A$, then $f(x)\in f(A)$, and by $(1)$, $x\in f^{-1}\left(f(A)\right)$. Therefore, by definition, we have $$ A\subset f^{-1}\left(f(A)\right)\tag{2} $$ However, if $f$ is not injective, then $f^{-1}\left(f(A)\right)$ may indeed contain elements not present in $A$; for example let $f:\mathbb{Z}\mapsto\mathbb{Z}$ be defined by $$ f(x)=\left\lfloor\frac x2\right\rfloor\tag{3} $$ and let $A$ be the set of even integers. Then $f(A)=\mathbb{Z}$ and $$ A\subsetneq\mathbb{Z}=f^{-1}\left(f(A)\right)\tag{4} $$

$\endgroup$
0
$\begingroup$

The problem can be reduced to one-line proof in the following way.

Let $f : Y \leftarrow X$ denote a function.

The definition of inverse images says that for all $y : Y$ and all $x : X,$ we have:

$$f^*(y) \ni x \iff y = f(x)$$

This can be rewritten:

$$f^*(y) \supseteq \{x\} \iff \{y\} \supseteq \{f(x)\}$$

This can be used to prove:

Proposition. For all subsets $B$ of $Y$ and all subsets $A$ of $X$, we have:

$$f^*(B) \supseteq A \iff B \supseteq f_*(A)$$

Once you've proved this, your problem becomes a one-line proof:

$$f^{*}(f_*(A)) \supseteq A \iff f_*(A) \supseteq f_*(A)$$

$\endgroup$

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.