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!

Is the following correct? (concerns sets and convergence)

  1. Dec 1, 2011 #1
    Let A = {(x,y) in R^2 | x^2 + y^2 <= 81}
    Let B = {((x,y) in R^2 | (x-10)^2 + (y-10)^2 <= 1}

    then here "A intersection B" is the empty set.
    Then let x_n be the sequence (0,10-(2/n)) which is a sequence in A and y_n be the sequence (10/n,10) which is a sequence in B.
    would |x_n - y_n| tend to 0 in this case?
  2. jcsd
  3. Dec 1, 2011 #2


    User Avatar
    Gold Member

    Why do you think xn is a sequence in A? It is not.
  4. Dec 1, 2011 #3
    Is it not? Isn't a sequence in A that converges to a point in A compliment?
  5. Dec 1, 2011 #4


    User Avatar
    Gold Member

    A sequence in A is a sequence all of whose values are in A. Almost all of the values of xn are not in A.
  6. Dec 1, 2011 #5
    A is a disk with radius 9. The sequence [itex]x_n[/itex] converges to (0,10). So the sequence is not in A.
  7. Dec 1, 2011 #6
    But can't a sequence x_n [itex]\subseteq[/itex] S converge to a point in S compliment?
  8. Dec 1, 2011 #7
    Not if S is closed, no.
  9. Dec 1, 2011 #8
    Oh right I see. Thank you.

    Can there exist a pair of non-empty closed sets A,B [itex]\subseteq[/itex] R^2 with A [itex]\cap[/itex] B equals the empty set with a pair of sequences x_n [itex]\subseteq[/itex] A and y_n [itex]\subseteq[/itex] B such that ||x_n - y_n|| -> 0.

    I was trying to find a specific example and hence my original thought was my first attempt.
  10. Dec 1, 2011 #9
    Yes, such a thing exists. Take

    [tex]A=\{(x,0)~\vert~x\in \mathbb{R}\}[/tex]


    [tex]B=\{(x,1/x)~\vert~x\in \mathbb{R}\}[/tex]

    (Note the distance between these two sets is 0). Can you find suitable sequences in these sets?
  11. Dec 1, 2011 #10
    Could you explain why the distance between the two sets is 0 please. I thought it was because d|x-y| = (x,0) - (x,1/x) = (0, -1/x) but wasn't sure how this means the distance is 0.

    Also I'm finding it hard to grasp the concept of sequences in 2 dimension in the sense of how to represent them but my guess would be,
    x_n = (3n, 0) and y_n = (3n, 1/3n) ?
  12. Dec 1, 2011 #11
    The distance between two sets is usually defined as the greater lower bound of the set of distances between 2 points in the sets. In symbols:
    d(A,B) = \inf \{d(a,b) : a \in A, b \in B\} \, .
    Take a look at this: http://en.wikipedia.org/wiki/Distance#Distances_between_sets_and_between_a_point_and_a_set

    So you were right that [itex] d(a,b) = |a-b| = |(x,0) - (x, 1/x)| = |(0,-1/x)| = 1/|x| [/itex]. But now you need to find the greatest lower bound of the set [itex] \{d(a,b) : a \in A, b \in B\} [/itex]. Do you see why it will be zero?
  13. Dec 2, 2011 #12
    Is it because as x gets smaller and tends to -infinity but you take the modulus and so u get limn-> infinity 1/x = 0?

    Also were my choices of sequences okay? That is the main issue I am stuck with? I need x_n in A and y_n in B such that ||x_n - y_n|| tends to 0.
    I think x_n=(3n,0) and y_n=(3n/1/3n) would work right?
  14. Dec 2, 2011 #13
    Yes, that choice of sequence is a good one!
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook