1
$\begingroup$

A homework problem guides me to prove that there is a rational number $\frac{m}{n}$ between every two real numbers $x$ and $y$. The first step requires me to prove that there exists an integer $n$ such that $\frac{1}{y-x} < n$. That you can always find a natural number larger than any given real number seems trivial obvious to me, but I can't figure out how to prove it. I thought of trying to use the fact that there is no largest natural number, but that doesn't prove that there is always one larger than a given real number. The assignment says that I can use the fact that $\frac{1}{y-x}$ has a decimal expansion to prove this first step, but I don't see how that helps.

Can someone please help me figure out how to prove that there exists an integer $n$ such that $\frac{1}{y-x} < n$ for all real $x$ and $y$?

$\endgroup$
  • $\begingroup$ Have you heard of the Archimedean Property? $\endgroup$ – AnotherJohnDoe Oct 1 '18 at 12:43
  • 5
    $\begingroup$ When needing to prove some "trivial" result you need to have a clear picture of what axioms you are beginning with. Your question appears to be lacking any statement about that. $\endgroup$ – user334732 Oct 1 '18 at 12:44
  • 1
    $\begingroup$ But between $\frac1{y-x}$ and $(\frac1{y-x})+1$ might be a logical place to look for such a number. $\endgroup$ – user334732 Oct 1 '18 at 12:47
2
$\begingroup$

Since $y-x>0$, by Archimedean property, there exits $n \in \Bbb{N}$ such that $n(y-x)>1$

[For, suppose it is false. Then we have $n \leq \frac{1}{y-x}, \forall n.$ which means $\frac{1}{y-x}$ is an upper bound for $\Bbb{N}$, a contradiction! ]

Question : Why we need that ?

Answer:

Assuming the existence of such an $\frac{m}{n}$ with $n > 0$. So, we have $x < \frac{m}{n}<y$. That is, $nx < m < ny$. Thus we are claiming that the interval $(nx, ny)$ contains an integer. It is geometrically obvious that a sufficient condition for an interval $J = (nx, ny)$ to have an integer in it is that its length $ny-nx$ is greater than $1$.That is , $n(y-x)>1$. Archimedean property assures of such $n$'s.

$\endgroup$
  • $\begingroup$ Thank you. Is there a way to use decimal expansion? $\endgroup$ – The Ledge Oct 1 '18 at 15:58

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.