How to find the distance between these elements?

  • Thread starter Thread starter futurebird
  • Start date Start date
  • Tags Tags
    Elements
futurebird
Messages
270
Reaction score
0
I'm studying the proof that the l2 metric space is complete and I just don't get something about this proof.

What I know:

  1. l2 is the metric space whose elements are sequences that converge when you square each term and sum them. In other words, a sequence, f is in l2 if SUM( |f(k)|2) < infinity.
  2. The norm on this metric space is: (SUM( |f(k)|2))0.5 = || f ||2
  3. A sequence in the l2 space is a sequence of sequences (confusing isn't it?)
  4. To show that the space is complete I need to show that every cauchy sequence in l2 converges to a sequence in l2.
My book says:

Suppose that for e > o there exists n0 such that:

||fn - fm||2 < e when m, n >= n0.

I don't understand how to think about ||fn - fm||2.

Suppose fn = .1, .01, .001, ...
Suppose fm = .2, .02, .002, ...

How would I find ||fn - fm||2 computationally?

Do I do:

||fn - fm||2 = (|(.12 + .012, .0012 + ...) - (.22 + .022 + .0022 + ...)|)0.5
||fn - fm||2 = (|(.01 + .001, .0001 + ...) - (.04+ .004 + .0004 + ...)|)0.5
||fn - fm||2 = (|(.01 + .001, .0001 + ...) - (.04 + .004 + .0004 + ...)|)0.5
||fn - fm||2 = ( |1/90 - 4/90|)0.5
||fn - fm||2 = (3/90)0.5

or

[(.1 - .2)2 + (.01-.02)2, + ...]0.5 ?

I think it's the first way... but for some reason I'm confused. Deeply.
 
Physics news on Phys.org
It's the second way. First find the difference between the sequences, then calculate the norm. Your first calculation is | ||fn|| - ||fm|| |.
 
Thanks dick... OK... So...

|| fn - fm||2 = [(.1 - .2)^2 + (.01-.02)^2 + ...]^0.5
|| fn - fm||2 = [(.1)^2 + (.01)^2 + ...]^0.5
|| fn - fm||2 = [.01 + .001 + ...]^0.5
|| fn - fm||2 = [1/90]^0.5

1 / sqrt(90) is about 0.105409255Grrr... Ok so:

|fn(k) - fm(k)| <= || fn - fm||2

|fn(k) - fm(k)| is the absolute difference between the kth terms in fn and fm ... sequences that are "close enough together" so that || fn - fm||2 < e.

Suppose

fn = .1, .01, .001, ...
fm = .2, .02, .002, ...
are that close together. This is saying that

.1 < = || fn - fm||2

.1 < = 0.105409255

Yay! it worked! I don't know why but it did!

Thanks!

God... but why is |fn(k) - fm(k)| <= || fn - fm||2?

I'm not quite seeing that...
 
Last edited:
Oh wait I think I see that now... it's like:

x-y <= sqrt((x-y)^2 + positive stuff or zero)I wish I knew what this was all about... It seems very disconnected from the other things in the course... I just don't know where it's going ...
 
futurebird said:
Oh wait I think I see that now... it's like:

x-y <= sqrt((x-y)^2 + positive stuff or zero)


I wish I knew what this was all about... It seems very disconnected from the other things in the course... I just don't know where it's going ...

You mean |x-y|. Yes, that's what it's like. BTW I get 1/sqrt(99) for your first example for ||fn-fm||. The common ratio in the geometric series is wrong. It's 1/100, not 1/10.
 
Dick said:
You mean |x-y|. Yes, that's what it's like. BTW I get 1/sqrt(99) for your first example for ||fn-fm||. The common ratio in the geometric series is wrong. It's 1/100, not 1/10.

ah yes.. see that now. OK... this proof is starting to make a little more sense to me now. Thank you!
 
There are two things I don't understand about this problem. First, when finding the nth root of a number, there should in theory be n solutions. However, the formula produces n+1 roots. Here is how. The first root is simply ##\left(r\right)^{\left(\frac{1}{n}\right)}##. Then you multiply this first root by n additional expressions given by the formula, as you go through k=0,1,...n-1. So you end up with n+1 roots, which cannot be correct. Let me illustrate what I mean. For this...
Back
Top