Prove l^p strict subspace of c0

  • Thread starter Thread starter looserlama
  • Start date Start date
  • Tags Tags
    Subspace
looserlama
Messages
29
Reaction score
0

Homework Statement



For F \in {R,C} and for an infinitie discrete time-domain T, show that lp(T;F) is a strict subspace of c0(T;F) for each p \in [1,∞). Does there exist f \in c0(T;F) such that f \notin lp(T;F) for every p \in [1,∞)

Homework Equations



Well we know from class that lp(T;F) = {f \in FT | Ʃ |f(t)|p < ∞} for p \in [1,∞) and c0(T;F) = {f \in FT | \forallε \in R>0, \exists a finite set S \subseteq T such that {t \in T | |f(t)| > ε} \subseteq S} (\Leftrightarrow t \rightarrow ±∞ \Rightarrow f(t) \rightarrow 0) are both vector spaces.

The Attempt at a Solution



Well as I said above, we know that both lp(T;F) and c0(T;F) are vector spaces, so to show that one is a subspace of the other it is sufficient to show that one is a subset of the other.

So,
Let f \in lp(T;F) \Rightarrow Ʃ|f(t)|p < ∞
Therefore lim as |t|\rightarrow∞ of |f(t)|p = 0 \Rightarrow lim as |t|\rightarrow∞ of |f(t)| = 0 \Rightarrow f \in c0(T;F)

Therefore lp(T;F) \subseteq c0(T;F).

Now this is the hard part, showing that it is a strict subset.

This is what I though of:

Define f \in FT by f(t) = \frac{1}{t<sup>1/p</sup>} if t ≠ 0 and 0 if t = 0.

Clearly f \in c0(T;F) as it's limit goes to 0.

And it's easy to show that f is not in any lp(T;F) for any specific p \in [1,∞).

But to properly do this problem I need to find a function that's not in lp(T;F) for every p \in [1,∞).

i.e., my problem is, I have to chose a p first, then this works, but whatever p I chose, f will not be in it's space, but it will be in the p+1 space. So it doesn't work for for every p, only a specific p.

So pretty much I can't think of a function that would be in c0 but not in lp for EVERY p \in [1,∞).

Any help would be awesome!
 
Physics news on Phys.org
Think about using logarithms.
 
Ok.

That makes sense as ln(t) increase much slower than t.

So this is what I was thinking:

Let f(t) = \frac{1}{ln(|t| + 1)} if t ≠ 0 and 0 if t = 0.

So clearly lim as |t|→∞ of f(t) = 0, so it is in c0.

But then showing it isn't in lp for every p is a bit harder.

This is how I tried:

Basically I wanted to use the comparison test to show Ʃ|\frac{1}{ln(|1| + 1)}|p = Ʃ1/|ln(|t| + 1)|p≤ Ʃ\frac{1}{|t|} and since that diverges \Rightarrow Ʃ|f(t)|p diverges.

The problem is showing that \exists N \in T such that \forall t ≥ N |ln(|t| + 1)|p ≤ |t|.

I tried differentiating both sides but it ends up giving p|ln(|t| + 1)|p - 1 ≤ |t| +1 which is essentially the same thing as before.

It makes sense to me as ln(t) increases much slower than t, and we can always find an N for which ln(t)p will be less than t for any t ≥ N. But I don't know how to show that.

I also thought of using the ratio or root test, but that seems like it wouldn't work very well...

Any thoughts?
 
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