0
$\begingroup$

I got stuck by proving following statement: $c \leq0 $ Prove: $ liminf_{n \to \infty }(ca_n) = c \, limsup_{n \to \infty }(a_n) $.

Prove:

Consider $A_n = \{a_m | m \geq n\} $. $A_n$ has a supremum. Let $x_n = sup(A_n)$. This applies that $\forall a \in A_n: x_n \geq a $. Next we multiply by $c \leq 0 $ and we get $ca \geq c x_n$. So $c \,x_n$ is a lower bound for the set $cA_n$. So we get: $ liminf_{n \to \infty }(ca_n) \leq c \, limsup_{n \to \infty }(a_n) $.

I got stuck by proving the other inequality. I tried it next way: Take $\epsilon \geq 0 $ random. Than $cx_n + \epsilon > cx_n$. So I just need to show that $cx_n + \epsilon$ is not a lowerbound for $cA_n$. But I can't get there. Somebody who wants to help me?

Thanks in advance!

$\endgroup$
1
$\begingroup$

Note that if $c \ge 0$ then $\sup_k c x_k = c \sup_k x_k$, and similarly for $\inf$.

Hence $\limsup_n c a_n = \lim_n \sup_{k \ge n} c a_k = \lim_n ( c \sup_{k \ge n} a_k )= c\lim_n \sup_{k \ge n} a_k = c \limsup_n a_n$, and similarly for $\inf$.

If $c \le 0$ instead, then we have $\sup_k c x_k = c \inf_k x_k$ and similarly with the $\sup, \inf$ interchanged. Repeating the above chain mutatis mutandis gives the desired result.

$\endgroup$
  • $\begingroup$ I suspect that $c\geq0$ in the first line of the question is a typo. The OP continues with: "....Next we multiply by $c\leq0$..." and I think that there $c\leq0$ is not a typo. $\endgroup$ – drhab Aug 28 at 13:51
  • $\begingroup$ Yeah, my mistake! Edited it! $\endgroup$ – Kabouter9 Aug 28 at 13:54
  • $\begingroup$ @drhab: Thanks for catching that. I suspect you are correct. $\endgroup$ – copper.hat Aug 28 at 13:54
  • $\begingroup$ I have added an answer for $c \le 0$. $\endgroup$ – copper.hat Aug 28 at 13:56

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.