Proving a subset F is closed if and only if any vector sequences in F's limit is also in F

28 Views Asked by At

I have a quick question regarding the "$\Leftarrow$" part of the proof, i'm not sure why something is true. It's a proof my professor write, and i'm trying to understand it. I'll just write out the proof until i get to my problem.

  • Assume if any vector sequence $ \in F$ converges, it's limit is also $\in F$.

  • Now i have to show $F$ is closed, or $F^c$ is open.

  • Let's have a look at any element in in $F^c $, call it $x$ so $x \in F^c$.

  • I have to find an $r>0$ such that $B(r,x) \subset F$

  • Let's assume $\nexists r$.

  • $\Rightarrow \exists B(r,x) \not\subset F^c$

  • $\Rightarrow \exists B(r,x) \cap F \not= Ø$
  • Since it's not empy, pick elements from there and call it $x^{(k)}$
  • We now have a vector sequence $ \{x^{(k)} \}^{\infty}_{k=1}$ where $x^{(k)} \in F \forall k \in \mathbb{N}$

And here is my problem. It supposbly should be true, that: $0 \leq x^{(k)} \in F \forall k $

Does anyone know why this inequality holds?

1

There are 1 best solutions below

2
On BEST ANSWER

The first 5 bullets are okay.

In fifth bullet is is assumed that so such $r>0$ exists.

That means that for every $r>0$ the set $B(r,x)\cap F$ is not empty.

Now e.g. for every $k>0$ pick some $x^{(k)}\in B(\frac1k,x)\cap F$.

Then $x^{(k)}\to x$ so that we may conclude that $x\in F$.

Now a contradiction has been reached and the conclusion is that the assumption that no such $r>0$ exists is wrong.

Then you are ready with proving that $F^{\complement}$ is open.

The inequality $0\leq x^{(k)}$ seems to "fall from the sky" and makes no sense at all in this context.