Trying to prove that a function converges at a point $x_0$ iff for every sequence $(a_n)_{n=0}^{\infty}$ with $a_n \to x_0$ also $f(a_n) \to L$

52 Views Asked by At

I am trying to prove this theorem from Analysis I by Tao.

Proposition 9.3.9. Let $X$ be a subset of $\mathbb{R}$, let $f : X → R$ be a function, let $E$ be a subset of $X$, let $x_0$ be an adherent point of $E$, and let $L$ be a real number. Then the following two statements are logically equivalent:

(a) $f$ converges to $L$ at $x_0$ in $E$. (i.e $\forall \varepsilon > 0 \exists \delta > 0 s.t \forall x \in E $ and $|x_{0}-x|\leq \delta , |f(x)-L|\leq \varepsilon$)
(b) For every sequence $(a_{n})_{n=0}^{\infty}$ which consists entirely of elements of $E$ and converges to $x_0$, the sequence $(f(a_{n}))_{n=0}^{\infty}$ converges to $L$.

I have (I think) been able to assume (a) and show that (b) follows. Let $(a_n)_{n=0}^{\infty}$ be a sequence converging to $L$.

Let $\varepsilon > 0$, and then I want to find an $N \in \mathbb{N}$ such that for every $n \geq N$

$$ |f(a_n) - L| < \varepsilon. $$

From the assumption in part (a) there must exist $\delta >0$ such that

$$ |a_n - x_0| < \delta \implies |f(a_n) - L| < \varepsilon $$

so if I can find $N \geq 1$ such that for all $n \geq N$ it is the case that $|a_n - x_0| < \delta$ then this direction is proven. By the fact that $a_n$ converges to $L$ for all $\varepsilon > 0$ (and thus also $\varepsilon = \delta$) this $N$ exists.

I am stuck on proving the other direction, assume (b) and attempt to prove (a).

Let $\varepsilon > 0$, and $\delta > 0$ (to be restricted later) and then $x \in E$ such that $|x - x_0| < \delta$.

Then I thought that I could use the fact that $x_0$ is an adherance point of $E$ to construct a sequence $(a_n)_{n=0}^{\infty}$. A very naïve guess is that because for every $\varepsilon > 0$ there exists a $y \in E$ such that $|y - x_0| < \varepsilon$, I could define the sequence $(a_n)_{n=0}^{\infty}$ such that $a_0$ is just the $x$ in question. Then let $a_1$ be such that $|a_1 - x_0| < \frac{\delta}{2}$, $a_2$ such that $|a_2 - x_0| < \frac{\delta}{3}$ and so on. Clearly this sequence converges to $x_0$, and therefore $(f(a_n))_{n=0}^{\infty}$ converges to $L$, by the assumption of (a).

However, in trying to exploit the fact that $(f(a_n))_{n=0}^{\infty}$ converges to $0$, I got a bit stuck because there are no guarantees that for a given $\varepsilon$, $f(a_0)$ is $\varepsilon$-close to zero (as the sequence must just be eventually $\varepsilon$-close).

I tried to think about what kind of sequence would be ideal - one which contains all the points in $E$ such that $|x - x_0| < \delta$ and converges to $L$, ideally (but I don't think that's possible as $E$ is not necessarily countable).

Sadly I'm kind of stuck here. Any hints/direction would be much appreciated. I think this should not be a very difficult proof and that I'm going about it the wrong way.

I thought maybe I should try to prove by contrapositive (but I have also seen advice not to do this as it is much easier to mistakes this way).

1

There are 1 best solutions below

0
On BEST ANSWER

(b) implies (a) is usually proved by contrapositive. If (a) does not hold, then there exists some $\varepsilon>0$ such that for every $\delta>0$ there exists some $x\in E$ for which $|x-x_0|<\delta$ but $|f(x)-L)|\geq\varepsilon$. In particular, for every $n$ there exists some $x_n\in E$ such that $|x_n-x_0|<\frac{1}{n}$ but $|f(x_n)-L|\geq\varepsilon$. Therefore there exists a sequence $x_n\in E$ such that $x_n\to x_0$ as $n\to\infty$ but $f(x_n)$ does not converge to $L$.