5
$\begingroup$

$\bf 3.7\ \ $ Theorem $\ \ $ The subsequential limits of a sequence $\{p_n\}$ in a metric space $X$ form a closed subset of $X$.

Proof $\ \ $ Let $E^*$ be the set of all subsequential limits of $\{p_n\}$ and let $q$ be a limit point of $E^*$. We have to show that $q\in E^*$.
$\qquad$ Choose $n_1$ so that $p_{n_1}\neq q$. (If no such $n_1$ exists, then $E^*$ has only one point, and there is nothing to prove.) Put $\delta=d(q,p_{n_1})$. Suppose $n_1,...,n_{i-1}$ are chosen. Since $q$ is a limit point of $E^*$, there is an $x\in E^*$ with $d(x,q)<2^{-i}\delta$. Since $x\in E^*$, there is an $n_i>n_{i-1}$ such that $d(x,p_{n_i})<2^{-i}\delta$. Thus $$d(q,p_{n_i})\leq 2^{1-i}\delta$$ for $i=1,2,3,...$. This says that $\{p_{n_i}\}$ converges to $q$. Hence $q\in E^*$.

Can someone explain how Rudin is selecting the $n_i$?

$\endgroup$

2 Answers 2

7
$\begingroup$

The sequence is choosen inductively. First he picks $n_1$ so that $p_{n_1}\neq q$ and set $\delta = d(p_{n_1}, q)$. Then he consider $\delta/2$. As $q$ is a limit point, there is $n_2$ (which can be chosen so that $n_2 > n_1$) so that

$$d(p_{n_2}, q) < \delta /2$$

Inductively, he choose $n_k>n_{k-1}> \cdots n_2 >n_1$ so that

$$d(p_{n_i}, q)< \frac{\delta}{2^i}.$$

$\endgroup$
1
  • $\begingroup$ I was wondering, why couldn’t we use the simpler proof: let $E^*$ be the set of subsequential limits. If this set has a limit point $q$ then for any $\epsilon>0$ there exists a subsequential limit $p \neq q$ such that $d(q, p) < \epsilon/2$ and there exists some K such that $d(p_{n_k}, p) < \epsilon/2$ for all $k\geq K$. Then for $k > K$, the distance $d(q, p_{n_k}) \leq d(q, p) + d(p, p_{n_k}) <\epsilon$ and so $q$ is a subsequential limit and thus in $E^*$? Is the problem that this cannot occur if $p\neq q$? $\endgroup$
    – akm
    Jul 19, 2020 at 18:27
2
$\begingroup$

If $E^* = \emptyset$, then $E^*$ is a closed subset of $X$.

If $E^* = \{q\}$ for some $q \in X$, then $E^*$ is a closed subset of $X$.

We assume that $\#E^* \geq 2$.

Let $q$ be a limit point of $E^*$.

If $p_n = q$ for all $n \in \{1,2,\cdots\}$, then any subsequence of $\{p_n\}$ converges to $q$. So $E^* = \{q\}$. But we assumed that $\#E^* \geq 2$. So there is $n_1 \in \{1,2,\cdots\}$ such that $p_{n_1} \neq q$.

Because $q$ is a limit point of $E^*$, there exists $x_2 \in E^*$ such that $d(x_2, q) < \frac{\delta}{2^2}$.

Because $x_2$ is a subsequential limit of $\{p_n\}$, there exists $n_2 \in \{1,2,\cdots\}$ such that $n_2 > n_1$ and $d(p_{n_2}, x_2) < \frac{\delta}{2^2}$.

By the triangle equality in $X$, $d(p_{n_2}, q) \leq d(p_{n_2}, x_2) + d(x_2, q) < \frac{\delta}{2^2} + \frac{\delta}{2^2} = \frac{\delta}{2}$.

Because $q$ is a limit point of $E^*$, there exists $x_3 \in E^*$ such that $d(x_3, q) < \frac{\delta}{2^3}$.

Because $x_3$ is a subsequential limit of $\{p_n\}$, there exists $n_3 \in \{1,2,\cdots\}$ such that $n_3 > n_2$ and $d(p_{n_3}, x_3) < \frac{\delta}{2^3}$.

By the triangle equality in $X$, $d(p_{n_3}, q) \leq d(p_{n_3}, x_3) + d(x_3, q) < \frac{\delta}{2^3} + \frac{\delta}{2^3} = \frac{\delta}{2^2}$.

$\cdots$

$\endgroup$

You must log in to answer this question.

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