Definition of a function by a metric.

Click For Summary

Discussion Overview

The discussion centers on the continuity of a function defined by a metric in a metric space, specifically exploring the function \( f_A(x) = p(x, A) \) where \( p \) is a metric on \( X \). Participants also delve into a separate question regarding the homeomorphism of a sphere minus a point to a Euclidean space.

Discussion Character

  • Technical explanation
  • Mathematical reasoning
  • Debate/contested

Main Points Raised

  • One participant suggests using the epsilon-delta definition of continuity to prove that \( f_A \) is continuous, highlighting the relationship \( p(x, A) \leq p(x, z) + p(z, A) \) for any \( z \) in \( X \).
  • Another participant agrees with the previous suggestion but emphasizes the need to apply it twice to obtain the absolute value on the left side, assuming familiarity with the epsilon-delta definition in metric spaces.
  • A separate question is posed about proving that \( S^n - \{x_0\} \) is homeomorphic to \( \mathbb{R}^{n+1} \), with a proposed approach involving dividing by the norm to maintain the unit norm of vectors.
  • One participant points out that the notation for the sphere is typically \( S^{n-1} \) and clarifies that removing a point from \( S^{n-1} \) makes it homeomorphic to \( \mathbb{R}^{n-1} \), not \( \mathbb{R}^{n+1} \), due to dimensional differences.
  • Another participant suggests using the function \( g(x) = x / ||x|| \) and notes the bijection between \( S^{n-1} \) and \( S^{n-1} - \{x_0\} \) to establish the homeomorphism.

Areas of Agreement / Disagreement

Participants express differing views on the dimensionality of the homeomorphic relationship, with some asserting it is to \( \mathbb{R}^{n-1} \) while others initially suggest \( \mathbb{R}^{n+1} \). The continuity of the function \( f_A \) remains under discussion without a clear consensus on the approach.

Contextual Notes

Participants rely on the epsilon-delta definition of continuity and the properties of metric spaces, but there are unresolved assumptions regarding the application of these concepts to the specific functions and spaces discussed.

MathematicalPhysicist
Science Advisor
Gold Member
Messages
4,662
Reaction score
372
Let (X,p) be a metric space with p metric on X, define for each subset A of X
p(x,A)=inf_{y\in A}p(x,y)
prove that: f_A(x)=p(x,A) f:X->R is continuous.
basically, if U is open in R, we need to show that f^-1_A(U) is open in X, i.e that
{x in X|p(x,A) in U} is open in X, now because U is open in R, then there's an open interval or half open interval contained in U that it contains p(x,A), i.e p(x,A) is bounded, but from this to show that f^-1(U) is open I'm kind of in a stuck.

any hints?

thanks in advance.
 
Physics news on Phys.org
It's probably easier to use the good old epsilon-delta definition of continuity here. The following observation will be helpful: p(x,A) <= p(x,z) + p(z,A), for any z in X.
 
Morphism is right. You have to do what he suggested twice, in order to get the absolute value on the left side. This is assuming that you've already learned that the epsilon-delta definition of continuity applies to metric spaces in general (and it does, calculus clases specialize in R^n, but it can be proven that it applies to metric spaces in general).
 
ok, I have another question, let S^n={x\in R^n|||x||=1}
let x_0=(1,0,...,0) (n entries), prove that S^n-{x_0} is homeomorphic to R^(n+1).
well I thought of dividing by the norm, i.e defining a homeomorphism such as this:
x=(x1,...,x_n+1) in R^n+1 f(x)=(x2,...,x_n-1,1)/||(x2,...,1)|| when (x1,...,xn)=(1,0,...,0)
when (x1,...,x_n)!=(1,0,...,0) then f(x)=(1,x2,...,x_n-1)/||(1,x2,...,x_n-1)||
I know that somewhere I need to divide by the norm in order to keep it that the norm of the vector is 1, but don'y know how, anyone has some hints on this?

thanks in advance.
 
hint: http://en.wikipedia.org/wiki/Stereographic_projection" .
btw, the set you defined as S^n is more usually written as S^{n-1}, and you need to show that removing a point makes it homeomorphic to R^{n-1}. It can't be homeomorphic to R^{n+1}, as they have different dimensions.
 
Last edited by a moderator:
so basically in order to prove it, all you need is to use g(x)=x/||x|| and the fact that S^n-1 is bijective to S^n-1\{x0}, so there's a bijection f between them then gof is the appropiate function.
 

Similar threads

  • · Replies 15 ·
Replies
15
Views
3K
  • · Replies 2 ·
Replies
2
Views
1K
  • · Replies 20 ·
Replies
20
Views
6K
  • · Replies 4 ·
Replies
4
Views
2K
  • · Replies 4 ·
Replies
4
Views
3K
  • · Replies 1 ·
Replies
1
Views
2K
Replies
9
Views
2K
  • · Replies 7 ·
Replies
7
Views
2K
  • · Replies 15 ·
Replies
15
Views
3K
  • · Replies 1 ·
Replies
1
Views
1K