1. Limited time only! Sign up for a free 30min personal tutor trial with Chegg Tutors
    Dismiss Notice
Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Proving the Greatest Lower Bound Property with

  1. Sep 13, 2011 #1
    1. The problem statement, all variables and given/known data
    Use part (a) to prove the Greatest Lower Bound Property.

    (a): If M is any upper bound for A, then: x[itex]\in[/itex](-A), -x[itex]\in[/itex]A, and -x[itex]\leq[/itex]M. Therefore x[itex]\geq[/itex]-M, hence -M is a lower bound for -A. By the Least Upper Bound Property, inf(-A) exists. If inf(-A) exists, then -M[itex]\leq[/itex]inf(-A) for all upper bounds M of the set A. Therefore, -sup(A)[itex]\leq[/itex]inf(-A).

    If -sup(A)<inf(-A), then it follows that there exists some distance between -sup(A) and inf(-A). Assuming set A is reflected correctly when transformed into set -A, it would be necessary for there to be a difference in the amount of elements in sets A and -A to create this distance. Set A must contain the same amount of elements as set -A by definition. Therefore, -sup(A) = inf(-A).

    Since A and B are bounded, [itex]\forall[/itex]a[itex]\in[/itex]A and [itex]\forall[/itex]b[itex]\in[/itex]B, there exist a1, a2[itex]\in[/itex]A and b1, b2[itex]\in[/itex]B such that a1[itex]\leq[/itex]a[itex]\leq[/itex]a2 and b1[itex]\leq[/itex]b[itex]\leq[/itex]b2 [itex]\forall[/itex]a[itex]\in[/itex]A and [itex]\forall[/itex]b[itex]\in[/itex]B. Thus (a1+b1)[itex]\leq[/itex](a+b)[itex]\leq[/itex](a2+b2) and (A+B) is also bounded.

    Let m=sup(A), n=sup(B). Then for all e>0,
    (i)x<m+(e/2) [itex]\forall[/itex]x[itex]\in[/itex]A and y<n+(e/2) [itex]\forall[/itex]y[itex]\in[/itex]B
    (ii)x>m-(e/2) for some x[itex]\in[/itex]A and y>n-(e/2) for some y[itex]\in[/itex]B

    [Combining the inequalities in each (i) and (ii)]

    (i) (x+y)<(m+n)+e [itex]\forall[/itex]x[itex]\in[/itex]A, [itex]\forall[/itex]y[itex]\in[/itex]B
    (ii) (x+y)>(m+n)-e for some x[itex]\in[/itex]A, for some y[itex]\in[/itex]B

    Hence, sup(A+B) = (m+n) = sup(A) + sup(B).


    2. Relevant equations
    See above.


    3. The attempt at a solution
    I haven't the slightest clue how to start this proof, so if you could give me that, I should be able to work out the rest of it.
     
  2. jcsd
  3. Sep 13, 2011 #2

    micromass

    User Avatar
    Staff Emeritus
    Science Advisor
    Education Advisor
    2016 Award

    You need to prove that each bounded-below set A has an infimum. Can you prove that -A has a supremum, first?
     
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook




Similar Discussions: Proving the Greatest Lower Bound Property with
Loading...