0

$f^\to$ is our image operator, i.e. $f^\to(A_\gamma) = \{f(x) \in Y: x \in A_\gamma\}$

I seek to prove that $f^\to$ preserves joins, i.e. $\forall \{A_\gamma: \gamma \in \Gamma\} \subset \wp (X)$, for any set $X$

$$f^\to (\displaystyle\bigcup_{\gamma \in \Gamma} A_\gamma) = \displaystyle\bigcup_{\gamma \in \Gamma} f^\to (A_\gamma)$$

I also seek to do this in two cases: $\Gamma = \varnothing$ and $\Gamma \neq \varnothing$


Proof

Case 1: $\Gamma \neq \varnothing$

I start this attempt by considering the RHS, i.e. $\displaystyle\bigcup_{\gamma \in \Gamma} f^\to(A_\gamma)$

By definition, for $f: X \to Y$ and $A_\gamma \subset X,$

$$f^\to(A_\gamma) = \{f(x) \in Y: x \in A_\gamma\}$$

Then, rewriting $\displaystyle\bigcup_{\gamma \in \Gamma} f^\to(A_\gamma)$, we have equivalently

$$\displaystyle\bigcup_{\gamma \in \Gamma} f^\to(A_\gamma)= \displaystyle\bigcup_{\gamma \in \Gamma} \{f(x) \in Y: x \in A_\gamma\}$$

Now, since $\Gamma \neq \varnothing, \exists \gamma \in \Gamma$ and thus

$$\gamma \in \Gamma \Rightarrow \exists A_\gamma \in \{A_\gamma: \gamma \in \Gamma\} \Rightarrow \{A_\gamma: \gamma \in \Gamma\} \neq \varnothing$$

Thus, $\displaystyle\bigcup f^\to(A_\gamma) \neq \displaystyle\bigcup \varnothing = \varnothing$

But from here, my confidence wavers: How might I proceed to show that the RHS is equivalent to the LHS, i.e. that $f^\to$ preserves joins?

Arctic Char
  • 16,007
SunRoad2
  • 661
  • You're just asking if images preserve unions? – Randall Sep 16 '21 at 14:00
  • @Randall Not quite, I'm attempting to prove so, actually – SunRoad2 Sep 16 '21 at 14:01
  • @Randall This link looks like it might be a good answer, provided that $f[A] = f^\to[A]$ since $f[A]$ is the image of a function and $f^\to[A]$ is the image operator. Can these be taken equivalently? – SunRoad2 Sep 16 '21 at 14:04
  • Why is it any different? Your question is all about sets, right? Then "join" is union and images are images. – Randall Sep 16 '21 at 14:05
  • @Randall I just wanted to make sure. Intuitively, I would agree that $f[A] = f^\to[A]$. But I had to clarify because we clearly have $f^\leftarrow[A] \neq f^{-1}[A]$, since $f^{-1}$ may not always exist (as an inverse function). I'm just being careful with my notation is all – SunRoad2 Sep 16 '21 at 14:09
  • 1
    Understood. This might be a case where using different notation for the same concept makes things more confusing (and unnecessarily so). – Randall Sep 16 '21 at 14:10

0 Answers0