Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Q is dense in R question about proof

  1. Aug 24, 2011 #1


    User Avatar


    In Rudin's Principles of Mathematical Analysis there's the following proof that Q is dense in R.

    Theorem: If [itex] x,y\in \mathbb{R}[/itex] and [itex]x < y [/itex] there exists a [itex]p \in \mathbb{Q} [/itex] such that [itex] x<p<y.[/itex]

    Proof: Since [itex]x<y[/itex], we have [itex]y-x>0[/itex]. It follow from the Archimedian property that there is a positive integer [itex]n[/itex] such that
    We again apply the Archimedian property to find positive integers [itex]m_1[/itex] a,d [itex]m_2[/itex] such that [itex]m_1>nx[/itex] and [itex]m_2>-nx[/itex]. Then
    Hence there is an integer [itex]m[/itex] (with [itex]-m_2\leq m\leq m_1[/itex]) such that
    [tex]m-1\leq nx < m.[/tex]

    We combine the inequalities to get
    [tex]nx < m \leq 1+nx < ny.[/tex]
    n is positive so
    [tex]x < \frac{m}{n} < y.[/tex]
    Which proves that [itex]\mathbb{Q}[/itex] is dense in [itex]\mathbb{R}[/itex].

    How one concludes that the m in the red bit exists is what's troubling me.

  2. jcsd
  3. Aug 24, 2011 #2
    Hmm, you're not the first one to ask that very question.

    Anyway, the key is to define

    [tex]m=\min\{k\in \mathbb{Z}~\vert~nx<k\}[/tex]

    However, there are some things we need to show:

    - Does the minimum exist?? I.e. is the set [itex]\{k\in \mathbb{Z}~\vert~nx<k\}[/itex] non-empty, does it have a lower bound? Why is the infimum a minimum?
    - does [itex]m-1\leq nx<m[/itex] hold.
  4. Aug 24, 2011 #3


    User Avatar
    Science Advisor
    Homework Helper

    i.e. are you aware of the well ordering principle?
  5. Aug 27, 2011 #4


    User Avatar

    Thanks for the replies, seems I don't know enough about integers yet, so I've been looking into them.
    I've found another book on analysis which does introduce the integers, it defines them as "those real numbers which are in every inductive set."

    But this isn't how the integers are usually introduced is it?

    What's a good book to learn enough about integers (and maybe a little more)?

  6. Aug 17, 2012 #5
    This is self evident .Any real number must lie between two successive integers . When You have the natural number system ,Irrationals fill the gaps between any 2 successive integers . It's very simple.
  7. Aug 17, 2012 #6
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook

Similar Threads - dense question proof Date
I Dense set equivalent definitions Feb 21, 2018
I Dense orbit and fixed point question Jan 2, 2017
Question on dense subset of l^p space Jul 16, 2015
Dense numbers Jul 4, 2015
Compact embedding and dense embedding Oct 31, 2013