0
$\begingroup$

I am trying to understand the proof of the theorem that if $f$ is continuous on a compact set $K \subseteq \mathbb{R}$ then $f$ is uniformly continuous on $K$. Here is the proof:

n

I am stuck on a couple things:

(1) How is $\lim [(y_{n_k})-(x_{n_k})] = 0$?

(2) The last statement of the proof claims that this proof has produced the desired contradiction. However, I don't understand how $\left| f(x_n) - f(y_n) \right| \geq \epsilon_0$ was contradicted by concluding that $\lim_{k \to \infty} \left| f(x_{n_k}) - f(y_{n_k}) \right| = 0$.

(3) [Edited from (2)] How does $\lim_{k \to \infty} \left| f(x_{n_k}) - f(y_{n_k}) \right| = 0$ imply $ \left| f(x_{n_k}) - f(y_{n_k}) \right| \geq \epsilon_0$ (in other words, where did the $\lim_{k \to \infty}$ part disappear)?

Any help is greatly appreciated!

$\endgroup$
3
  • $\begingroup$ 1. In the beginning it is assumed, $|x_n-y_n| \to 0$, and $ x_{n_k} - y_{n_k} $ is a sub-sequence of that sequence. $\endgroup$ Commented May 8, 2020 at 6:09
  • 1
    $\begingroup$ 3. It does not imply it, it contradicts it by the definition of limit. $\endgroup$ Commented May 8, 2020 at 6:20
  • $\begingroup$ If something always remains greater than (or equal to) a specific positive number $\epsilon_0$ then that something cannot tend to $0$. This obvious fact explains your 2). $\endgroup$
    – Paramanand Singh
    Commented May 8, 2020 at 6:24

2 Answers 2

1
$\begingroup$

First question: If a sequence tends to $0$ so does every subsequence. Since $y_n-x_n \to 0$ so does $y_{n_k}-x_{n_k}$.

Second question: If $|c_n| \geq \epsilon_0$ for all $n$ then $|c_{n_k}| \geq \epsilon_0$ for all $k$. This implies that $c_{n_k}$ cannot tend to $0$.

$\endgroup$
3
  • $\begingroup$ Thanks for the clear answer! Based on your answer for (2), I interpreted: $\lim [f(x_{n_k})-f(y_{n_k})] = 0$ implies $\forall \epsilon > 0, \exists N \in \mathbb{N}$ s.t. $k \geq N \implies \left| f(x_{n_k})-f(y_{n_k}) \right| < \epsilon$. Now, since this implication is true from all $k \geq N$, we have $\left| f(x_{n})-f(y_{n}) \right| < \epsilon$ which contradicts $\left| f(x_{n})-f(y_{n}) \right| \leq \epsilon_0$. Is this interpretation correct? $\endgroup$ Commented May 8, 2020 at 19:13
  • $\begingroup$ @Ricky_Nelson It is correct but there is a typo in your last inequality. $\endgroup$ Commented May 8, 2020 at 23:30
  • $\begingroup$ I see, the last equality should read $\geq \epsilon_0$. Thanks you very much for your answer! $\endgroup$ Commented May 9, 2020 at 0:52
0
$\begingroup$

According to the theorem as $ \lim_{n\to\infty} |x_n -y_n| \to 0$ we have $ |f(x_n)-f(y_n)| \to 0 $ the proof above considers the first part to be true and by contradiction shows that the second part has to be true in case of a function defined on a compact set. As both $x_n$ and $y_n$ converge to $x$ we must have by the Algebraic Limit Theorem that $f(x_n)%$ and $f(y_n)$ converges to $f(x)$. Thus there exist an $n > N_0$ such that $|f(x_n)-f(y_n)|<\epsilon$ , hence the contradiction.

$\endgroup$

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .