MHB Can you prove Tate's theorem on bounded mappings?

  • Thread starter Thread starter Euge
  • Start date Start date
  • Tags Tags
    2016
Click For Summary
The discussion revolves around proving Tate's theorem, which states that if a mapping $\phi : X \to Y$ of real Banach spaces satisfies a bounded condition on the difference between $\phi(x + y)$ and the sum of $\phi(x)$ and $\phi(y)$, then there exists a unique additive mapping $\psi : X \to Y$ where the difference $\psi - \phi$ is bounded. The problem of the week (POTW) did not receive any responses, indicating a lack of engagement or difficulty in solving it. A solution to the theorem is provided by the original poster. The discussion emphasizes the importance of understanding bounded mappings in the context of functional analysis. The thread highlights the challenge of proving such theorems in mathematical discussions.
Euge
Gold Member
MHB
POTW Director
Messages
2,072
Reaction score
245
Here is this week's POTW:

-----
Prove the following theorem of Tate: If $\phi : X\to Y$ is a mapping of real Banach spaces such that for some positive number $M$, $\lvert \phi(x + y) - \phi(x) - \phi(y)\rvert \le M$ ($x,y\in X$), then there is a unique additive mapping $\psi : X \to Y$ such that $\psi - \phi$ is bounded on $X$.

-----
Remember to read the http://www.mathhelpboards.com/showthread.php?772-Problem-of-the-Week-%28POTW%29-Procedure-and-Guidelines to find out how to http://www.mathhelpboards.com/forms.php?do=form&fid=2!
 
Physics news on Phys.org
No one answered this week's problem. You can read my solution below.
Note $\phi(2x) = 2\phi(x) + O(x)$, where $\lvert O(x) \lvert < M$. Inductively, $\phi(2^nx) = 2^n\phi(x) + (2^n-1)O(x)$ for all positive integers $n$. Fix $x\in X$ and consider the sequence $\phi_n := \frac{\phi(2^nx)}{2^n}$, $n\in \Bbb N$. Since

$$|\phi_n - \phi_m| = \left\lvert \left[\phi(x) + \left(1 - \frac{1}{2^n}\right)O(x)\right] - \left[\phi(x) + \left(1 - \frac{1}{2^m}\right)O(x)\right] \right\rvert = \left\lvert \left(\frac{1}{2^m} - \frac{1}{2^n}\right)O(x)\right\rvert = \left(\frac{1}{2^m} - \frac{1}{2^n}\right)O(x) \to 0$$

as $n,m\to \infty$, then $\phi_n$ is a Cauchy sequence in $Y$. Since $Y$ is complete and $x$ was arbitrary, there is a mapping $\psi : X\to Y$ such that $\lim\limits_{n\to \infty} \phi_n(x) = \psi(x)$ for all $x\in X$. Now

$$\psi(x + y) = \lim_{n\to \infty} \frac{\phi(2^nx + 2^n y)}{2^n} = \lim_{n\to \infty} \frac{\phi(2^nx) + \phi(2^ny) + E(x, y)}{2^n} = \psi(x) + \psi(y) + \lim_{n\to \infty} \frac{E(x,y)}{2^n},$$

where $\lvert E(x,y)\rvert < M$. Thus $\lim\limits_{n\to \infty} \frac{E(x,y)}{2^n} = 0$ and $\psi$ is additive.

To see that $\psi - \phi$ is bounded, fix $x\in X$, and choose a positive integer $N = N(x)$ such that $\lvert \psi(x) - \phi_N(x)\rvert < M$. Then

$$\lvert \psi(x) - \phi(x)\rvert \le \lvert \psi(x) - \phi_N(x)\rvert + \lvert \phi_N(x) - \phi(x)\rvert < M + M = 2M$$

Since $x$ was arbitrary, $\psi - \phi$ is bounded on $X$.

For uniqueness, let $\theta$ be any additive mapping from $X$ to $Y$ such that $\theta - \phi$ is bounded on $X$. Given $x\in X$, $\theta(x) = \frac{\theta(2^n x)}{2^n}$ for all $n\in \Bbb N$. Thus

$$\lvert \theta(x) - \phi_n(x)\rvert = \left\lvert \frac{\theta(2^nx) - \phi(2^nx)}{2^n}\right\rvert \le \frac{C}{2^n},$$

where $C$ is a constant. Hence

$$\theta(x) = \lim_{n\to \infty} \phi_n(x) = \psi(x)$$
 

Similar threads

Replies
1
Views
2K
Replies
1
Views
2K
  • · Replies 1 ·
Replies
1
Views
2K
  • · Replies 1 ·
Replies
1
Views
2K
  • · Replies 1 ·
Replies
1
Views
3K
Replies
1
Views
2K
Replies
1
Views
3K
  • · Replies 1 ·
Replies
1
Views
3K
  • · Replies 1 ·
Replies
1
Views
3K
  • · Replies 2 ·
Replies
2
Views
3K