Isomorphic Hilbert spaces iff they have the same dimension

942 Views Asked by At

This is a specific question about the proof of Theorem 5.4 in "A Course in Functional Analysis" by Conway. The theorem is stated in the question.

enter image description here

The highlighted line confuses me. I thought it should be "ran U contains all functions f in l^2(E) such that f(e) = 0 for all but a COUNTABLE number of e". By Corollary 4.9 in the same book, if $E$ is an orthonormal set in $H$ and $h \in H$, then $<h, e> \neq 0$ for at most a countable number of vectors $e$ in $E$. If $H$ is infinite-dimensional (so its basis $E$ will also has infinite number of elements), and we pick $h$ as any element in its basis $E$, then we would have $f(e) := <h,e> = 0$ for all $e \neq h$ in $H$, i.e. infinitely many $e$'s.

1

There are 1 best solutions below

0
On

The set A containing all functions f such that f(e) = 0 for all but a FINITE number of e is contained in the set B of "... but a COUNTABLE number ...", hence it is no contradiction to Corallary 4.9.

I think the author likes to use the "smaller" basis A, which is sufficient for his argumentation, but you could do the same argumentation with set B instead of A, as both A and B are dense in l^2(E).