Let $H$ be a Hilbert space and $T \in \mathcal{L}(H)$ be self-adjoint. I know that this is a basic question, but I do not understand the infinite dimensional case of linear algebra very well. My question is, when does there exist a Hilbert basis for $H$ consisting entirely of eigenvectors of $T$? In the finite dimensional case, we only had to show that the matrix representation of $T$ in any basis was similar to a diagonal matrix. But in the infinite dimensional case, it is not clear that matrix representations even make any sense.

  1. Example: let $H=L^2[0,1]$ and let $T:H \to H$ be defined by $(Tf)(t)=tf(t)$. Then $T$ is self-adjoint but has no eigenvalues.

  2. If $A:H \to H$ is compact and self-adjoint, then there exists an orthonormal basis of $H$ consisting of eigenvectors of $A$. (See: https://en.wikipedia.org/wiki/Compact_operator_on_Hilbert_space).

  • $\begingroup$ Ah, thank you! I was looking for a characterization like this. $\endgroup$
    – J_Psi
    Nov 30 '18 at 9:12

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.