1. Not finding help here? Sign up for a free 30min 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!

Metric space completion

  1. May 1, 2007 #1
    1. The problem statement, all variables and given/known data

    Let X be an incomplete metric space. and Let X' denote its completion. I would like to show that there is Cauchy sequence in X which does not converge in X but does converge in X'. Moreover, I want to show that X contains every element of the sequence except the limit point.

    2. Relevant equations

    No equations as such other than the

    Definition: A complete metric space is one in which every Cauchy sequence converges.

    Definition: A Cauchy sequence { x_n } is one for which given any epsilon > 0 there exists a natural N such that if m,n > N then d(x_n,x_m) < epsilon.

    3. The attempt at a solution

    Well, negating the definition of a metric space we should be able to find a cauchy sequence which does not converge in X.

    My next line of argument would be to say that since X is isometrically embedded in X'. Then { x_n } is also a Cauchy sequence in X' and thus converges in X'.

    However, I'm not sure if it is always the case that X' \ X will always only contain the limit point. Can it not contain other points in the sequence?
    Last edited: May 1, 2007
  2. jcsd
  3. May 1, 2007 #2


    User Avatar
    Staff Emeritus
    Science Advisor

    I assume that was a typo and you mean "negating the definition of complete". Yes, since you are given that X is NOT complete, there must exist a Cauchy sequence in X which does not converge.

    Technically, there exist an isometric ((distance preserving), one-to-one function , f(xn)= x'n is in X'. Can you determine whether x'n is a Cauchy sequence in X'? If so, then since you know that X' is complete, the sequence {x'n} must converg to some point x'. Does there exist x in X such that f(x)= x'? If so what would be its relation to the sequence {xn}?
  4. May 1, 2007 #3
    Yes, we just choose the same natural number N since d(x_n,x_m) = d'(x_n',x_m').

    If there existed x in X such that f(x) = x'. Then d'(x', x_m') = d(x, x_m). But lim d(x,x_m) = lim d'(x',x_m') = 0 which would imply that { x_n } converges in X (contradiction).

    I'm still struggling to show that f(X) contains every element of the sequence except the limit point. Is this actually true?
  5. May 2, 2007 #4


    User Avatar
    Staff Emeritus
    Science Advisor

    What, exactly, is the definition of "completion" of a metric space?
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook

Have something to add?

Similar Discussions: Metric space completion
  1. COMPLETE metric space (Replies: 28)