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!

Homework Help: Closed Sets

  1. Dec 1, 2009 #1


    User Avatar
    Gold Member

    1. The problem statement, all variables and given/known data

    Prove that a closed rectangle [itex]A \subset \mathbb{R}^n[/itex] is a closed set.

    2. Relevant equations


    3. The attempt at a solution

    Let [itex]A = [a_1,b_1] \times \dots \times [a_n,b_n] \subset \mathbb{R}^n[/itex], then [itex]A[/itex] is closed if and only if its complement, [itex]\mathbb{R}^n - A[/itex], is open. Now, let [itex]x = (x_1, \dots, x_n) \in \mathbb{R}^n - A[/itex] and choose [itex]\varepsilon_i = \inf_{c_i \in [a_i,b_i]}\frac{|x_i - c_i|}{2} > 0[/itex]. Clearly we have that [itex](x_i - \varepsilon_i,x_i + \varepsilon_i) \cap [a_i,b_i] = \emptyset[/itex]; thus, we can define an open rectangle [itex]B = (x_1 - \varepsilon_1, x_1 + \varepsilon_1) \times \dots \times (x_n - \varepsilon_n, x_n + \varepsilon_n)[/itex] [itex]\subset \mathbb{R}^n - A[/itex] with the property that for any [itex]x \in \mathbb{R}^n - A[/itex], [itex]x \in B \subset \mathbb{R}^n - A[/itex], completing the proof.

    My questions:
    1) Is this "proof" correct?
    2) If so, is there an easier way to do this?

  2. jcsd
  3. Dec 1, 2009 #2
    No. [tex]\varepsilon_i[/tex] may be zero for some i. Consider the unit square [0,1] x [0,1] in [tex]\mathbb{R}^2[/tex]. The point (1/2, 2) is not in the unit square, but your [tex]\varepsilon_1=0[/tex].

    An easier proof: recall that a set is closed if and only if for a sequence [tex]x^{(k)}[/tex] in the set that converges, the limit is in the set. Now let [tex]x^{(k)}[/tex] be a sequence in A that converges to the point x. For i = 1, 2,..., n we have [tex]x_i^{(k)} \to x_i[/tex] in R. Now, [tex][a_i,b_i][/tex] is closed in R (this is very easy to prove), so [tex]x_i \in [a_i,b_i][/tex] which impies that [tex]x \in A[/tex].
  4. Dec 1, 2009 #3


    User Avatar
    Gold Member

    Ah! That's a really good point. :grumbles quitely:

    I haven't learned that formulation of a closed set so I can't really follow your proof (sorry!). So, suppose I revised my proof as follows: Let [itex]x = (x_1, \dots, x_n) \in \mathbb{R}^n - A[/itex], then for some [itex]x_i[/itex] we have that [itex]x_i \notin [a_i,b_i][/itex]. I can then choose [itex]\varepsilon = \inf_{c_i \in [a_i,b_i]}\frac{|x_i - c_i|}{2} > 0[/itex] and since [itex](x_i - \varepsilon, x_i + \varepsilon) \cap [a_i,b_i] = \emptyset[/itex] I can define the open rectangle [itex]B = (x_1 - \varepsilon, x_1 + \varepsilon) \times \dots \times (x_n - \varepsilon, x_n + \varepsilon) \subset \mathbb{R}^n - A[/itex]. Hence, for any [itex]x \in \martbb{R}^n - A[/itex] there exists an open rectangle [itex]B[/itex] such that [itex]x \in B \subset \mathbb{R}^n - A[/itex], completing the proof.

    Does this work any better? I just started learning about open and closed sets today so I appreciate all the help that I can get. Thanks again!
  5. Dec 1, 2009 #4
    This almost works like a charm. But the fact that [tex]\varepsilon > 0[/tex] requires you to prove that [tex][a_i,b_i][/tex] is closed in R. There's an easier way to pick [tex]\varepsilon = \frac{ 1 }{ 2 } \min \{ |x_i-a_i|, |x_i-b_i| \}[/tex].
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook