There are a bunch of questions I've gone through here that talk about how to solve a specific question. I am more curious on the intuition behind a particular problem. I am self studying probability theory, and am now in the analysis section.
I understand the general concepts of infinum and supremum:
$\inf A_k = \cap_{k=n}^\infty A_k$
The intuition here is that the intersection of all possible of k-length subsequences will be the greatest lower bound.
The same goes for supremum:
$\sup A_k = \cup_{k=n}^\infty A_k$
In this case the union of all possible k-length subsequences will be the least upper bound.
This problem however has me confused:
Check that $\liminf_{n \to \infty} [0, \frac{n}{n+1}) = \limsup_{n \to \infty}[0, \frac{n}{n+1}) = [0, 1)$
Setting out to try this I attempted to solve one part of it:
$\liminf_{n \to \infty} [0, \frac{n}{n+1})$
$\lim_{n \to \infty} [\inf_{k \le n} [0, \frac{n}{n+1})]$
$\cup_{n = 1}^\infty \cap_{k=n}^\infty [0, \frac{n}{n+1})$
So I need to try to find the union of all intersections of the subsequences. This is where my intuition falls apart. Writing down some test sequences:
$k = 1: [0, \frac{1}{2}) \cap [0, \frac{2}{3})\cap [0, \frac{3}{4}) \cap ...$
$k = 2: [0, \frac{2}{3}) \cap [0, \frac{3}{4}) \cap [0, \frac{4}{5}) \cap ...$
So now looking a bit deeper at $k = 1$
$[0, \frac{1}{2}) \cap [0, \frac{2}{3})$
Would give me $[0, \frac{1}{2})$ since its the only common subsequence right? Continuing down the line, it would appear eventually this repeated intersectioning would result in $[0, 1)$. This would mean that unioning all of the results of the intersectioning would also be $[0, 1)$ - right? Doing this for $k = 2$ reveals the same thing.
Am I proceeding through this question correctly? I don't have a professor available to ask...so I want to take the time to make sure I am doing this correctly.
Thank you!
EDIT: I've realized part of my interpretation was wrong:
For
$k = 1: [0, \frac{1}{2}) \cap [0, \frac{2}{3})\cap [0, \frac{3}{4}) \cap ...$
This intersection will only ever result in $[0, \frac{1}{2})$ because it is all the sequences here have in common. For $k = 2$ this value will be $[0, \frac{2}{3})$. This continues on. Then, when you get to the unioning part you end up with:
$[0, \frac{1}{2}) \cup [0, \frac{2}{3}) \cup ...$
Which is the union of all of the k intersection results up to infinity.
etc etc. If you do this, it is obvious the union becomes $[0, 1)$ in the limit.
Is this more correct?