3
$\begingroup$

I am taking a self study graduate course on Functional Analysis and the notes are quite terse. I am struggling in particular with the linear algebra background part, including dealing with infinite dimensional vector spaces.

So I was looking for a good reference to complement the notes. It would be particularly good if the book involved solved exercises and detailed proofs.

Thank you.

$\endgroup$
  • $\begingroup$ Any good, undergraduate level introductions to Functional Analysis? might not be a duplicate, but might still be helpful. $\endgroup$ – Mark S. Oct 28 at 13:08
  • $\begingroup$ @Mark Thanks for your input but I already saw thst wuestion amd run into the same problems than with other references: it covers topology and banach and hilbert spaces but not arbitrary infinite dimensional vector spaces and isomorphisms etc... But thank you anyways. $\endgroup$ – Student Oct 28 at 14:20
2
$\begingroup$

I'm not aware of any such reference that isn't either a text on functional analysis or a text on commutative algebra. It would help if you were more specific about what you wanted to know about infinite-dimensional vector spaces. Most of what is worth knowing about them is a list of things that don't generalize from finite-dimensional vector spaces. If $V$ is infinite-dimensional:

  • In general you can't define the determinant, trace, or characteristic polynomial of an endomorphism $T : V \to V$.
  • It's not true that $T : V \to V$ must have an eigenvalue even over an algebraically closed field.
  • It's not true that $V$ has the same dimension as $V^{\ast}$.
  • It's not true that an injection $V \to V$ is automatically a bijection, and similarly for surjections.

Then there's some stuff that relies on the axiom of choice:

  • $V$ still has a basis, but only if you assume the axiom of choice (and in fact it's a famous result due to Blass that AC is equivalent to the assertion that every vector space has a basis). Due to the independence of AC it follows that it's consistent with ZF that there are infinite-dimensional vector spaces without bases.
  • Linear functionals still separate points, but again only if you assume the axiom of choice. It's consistent with ZF that there are infinite-dimensional vector spaces with no nonzero linear functionals (see this MO thread).
| cite | improve this answer | |
$\endgroup$
  • $\begingroup$ Thank you. I am not looking for commutative algebra. Do you have any refereces of texts on functional analysis that treat infinite dimensional vector spaces in generality? For example, I would like to see a treatment of basis for infinite dimensional spaces witb maybe some examples and I would also like to see the proof for infinite dimensional spaces rhat two vector spaces are isomorphic if and only if any of their basis has the same cardinality, a proof that the basis for thr dual space of V has cardinality 2^card(B) for a basis B of V. This sort of thing. $\endgroup$ – Student Oct 29 at 22:26
  • $\begingroup$ @Student: I'm not aware of a reference for any of that. I would consider all of those basically annoying exercises. Many infinite-dimensional vector spaces don't have explicit bases in any reasonable sense so it's unclear what "some examples" would mean here. $\endgroup$ – Qiaochu Yuan Oct 29 at 22:36
  • $\begingroup$ The way the proofs go for bases in the presence of choice is a Zorn's lemma argument: you start with a linearly independent set of vectors and just keep adding more until you get a maximal linearly independent set, which must be a basis. For a proof that two bases have the same size see Wikipedia: en.wikipedia.org/wiki/Dimension_theorem_for_vector_spaces#Proof $\endgroup$ – Qiaochu Yuan Oct 29 at 22:39

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.