1

Let $(r_{n})$ be an enumeration of the set $\mathbb{Q}$ of all rational numbers. Show that there exists a subsequence $(r_{n_{k}})$ such that $\lim_{k \to \infty} = +\infty$.

Theorem If the sequence $(s_{n})$ is unbounded above, it has a subsequence with limit $+\infty$.

Please critique or vote on the proof I supply as a response. Thanks!

skm
  • 81

1 Answers1

0

Proof. Since $\mathbb{Q}$ is unbounded above, by the theorem above, we know that $(r_{n})$ contains a subsequence that diverges to $+\infty$.

Lemma: $\mathbb{Q}$ does not have an upper bound.

Proof of Lemma. Suppose for contradiction that $\mathbb{Q}$ does have an upper bound. That is, there is some $q$ such that for all $x \in \mathbb{Q}$, we have $q \geq x$.

First consider some positive $z \in \mathbb{Q}$, and fix some corresponding $r \in \mathbb{Q}$ such that we know $z \geq \frac{q}{r}$. Notice here that $q$ must be positive since 100000000 is a member of $\mathbb{Q}$ and $q$ must be greater than it.

We have two cases to consider: either $r$ is positive or $r$ is negative. Suppose $r$ is positive; we know that $zr \geq q$ which is a contradiction since $zr \in \mathbb{Q}$ due to multiplicative closure.

Suppose there is no positive $r$ such that $z \geq \frac{q}{r}$. That is for all $r>0$, $z < \frac{q}{r}$. Consider then $r=-1$ in which case $z > q$ contradicting our original claim as required.

skm
  • 81
  • "we can choose some arbitrarily large integer" means you already know $\mathbb{Q}$ is unbounded above –  Jul 24 '19 at 00:03
  • Ahh. I'll change that. – skm Jul 24 '19 at 05:08
  • " That is for all $r>0$, $z < \frac{q}{r}$." assumes $r>0$, but then you use $r=-1<0$. –  Jul 24 '19 at 09:38
  • Actually why you don't just consider q+1? –  Jul 24 '19 at 09:38
  • If I'm reading your statement right, q+1 might not work. q doesn't necessarily have to rational, and the closure property wouldn't apply. – skm Jul 25 '19 at 03:28
  • You can take the integer part of q+1, which is obviously larger than q –  Jul 25 '19 at 07:40
  • Yup, that would work too. It's the same as choosing some integer below $q$ though it is simpler to write that than how I do it. – skm Jul 29 '19 at 14:33