Uniform Continuity of Sequences in Metric Space

Join the discussion
Registration is free. Ask a follow-up in this thread, or start your own.
1 reply · 3K views
jdcasey9
Messages
25
Reaction score
0

Homework Statement


Prove that f:(M,d) -> (N,p) is uniformly continuous if and only if p(f(xn), f(yn)) -> 0 for any pair of sequences (xn) and (yn) in M satisfying d(xn, yn) -> 0.



Homework Equations






The Attempt at a Solution


First, let f:(M,d)->(N,p) be uniformly continuous.

Let [tex]\epsilon[/tex]=2[tex]\delta[/tex].

lf(xn)-f(yn)l [tex]\leq[/tex] lf(xn)-xnl + lxn-f(yn)l [tex]\leq[/tex] lf(xn)-xnl + lxn-ynl + lyn-f(yn)l < [tex]\delta[/tex] + 0 + [tex]\delta[/tex]= 2[tex]\delta[/tex] =[tex]\epsilon[/tex]
(because f is uniformly continuous)

Therefore, p(f(xn), f(yn))->0.

Second, let p(f(xn), f(yn)) -> 0 for (xn), (yn) in M such that d(xn, yn) ->0.

We can do this nearly the same way, except at the end we say:

lf(xn)-f(yn)l [tex]\leq[/tex] lf(xn)-xnl + lxn-ynl + lyn-f(yn)l -> 0 so it must be uniformly continuous.
 
Physics news on Phys.org
jdcasey9 said:
lf(xn)-f(yn)l [tex]\leq[/tex] lf(xn)-xnl + lxn-f(yn)l [tex]\leq[/tex] lf(xn)-xnl + lxn-ynl + lyn-f(yn)l < [tex]\delta[/tex] + 0 + [tex]\delta[/tex]= 2[tex]\delta[/tex] =[tex]\epsilon[/tex]

Firstly, how did you define the absolute value? Absolute value is only defined on R, but now you're working in an arbitrary metric space.
Secondly, how did you define f(x)-f(y). Again, you're working in an arbitrary metric space, thus it may be that there is no addition/substraction defined on that space.