I am interested in finding literature pertaining the problem posed by this question, which is the degree to which an operator $A$ on an infinite dimensional (separable) Hilbert $X$ space can be "approximated" by matrices $A_n$ corresponding to the restriction of $A$ to the $n$th member of an ascending chain of subspaces $U_1\subset U_2 \subset \cdots \subset U_n \subset \cdots \subset X$ for which

(1) $\dim(U_n) = n$, and

(2) $\lim_{n\rightarrow \infty} U_n = X.$

It is mentioned that such an "approximation" theory was developed around the 1900s, but I have no functional analysis text which mentions any such development. Specific questions of interest would be

(i) Let $\sigma(P)$ be the spectrum of $P$. When must we have $\lim_{n\rightarrow \infty}\sigma(A_n) = \sigma(A)$? (a statement is made without proof that this is true for compact operators).

(ii) Is there ever a case for which $\lim_{n\rightarrow \infty}\|A_n\|_{op} = A$? (another statement is made that there are no such cases).

(iii) Let $x \in X$ and suppose $x_n$ is its restriction to $U_n$. When is it true (if ever) that $\lim_{n\rightarrow\infty} A_nx_n = Ax$? (this may just be a restatement of (ii).)

Does anyone know of a good standard reference here? A text or detailed review article would be the best case.

  • 2
    $\begingroup$ More recent than the book mentioned by Piyush Grover: I think Roch, Silbermann and others have investigated conditions under which one gets better behaviour of "finite approximation methods". See also arxiv.org/abs/1005.0166 by Chandler-Wilde and Lindner $\endgroup$ – Yemon Choi Jan 3 '18 at 19:11
  • $\begingroup$ @YemonChoi what would you recommend by Roch and/or Silbermann? $\endgroup$ – SZN Jan 3 '18 at 19:14
  • $\begingroup$ The book I am aware of (but which I do not own, and which I have not read thoroughly) is books.google.co.uk/books?id=CoT1BwAAQBAJ The monograph that I linked to in my previous comment may be more up to date $\endgroup$ – Yemon Choi Jan 3 '18 at 19:56

A canonical reference is : Spectral Approximation of Linear Operators by Chatelin.


| cite | improve this answer | |

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.