4
$\begingroup$

Let $H$ be a Hilbert space, $A$ be a normal bounded operator on $H$ with spectrum $\sigma(A)=\{\lambda\in \mathbb{C}\;|\;A-\lambda Id \text{ is not invertible }\}$. Is $\sigma\left(\dfrac{A-A^*}{2i}\right)$ the set of imaginary part of the elements of $\sigma(A)$ ? Thanks.

$\endgroup$
7
$\begingroup$

There is probably an elementary proof, but that's an immediate consequence of the continuous functional calculus :

$ \frac{A-A^*}{2i} = f (A) $

where $f$ is the imaginary part function on $\mathbb{C}$. And when you apply a continuous function $f$ to a normal operator $A$ you have : $Spec(f(A)) = f(Spec(A))$ (you can see that by restricting to the abelian sub-$C^*$-algebra generated by $A$).

| cite | improve this answer | |
$\endgroup$
0
$\begingroup$

The spectral theorme assert that there exist a unitary operator $U$ sucht that $U^{*}AU$ is the multiplication operator by $\lambda$, $\lambda \in \mathbb{C}$. With the same unitary opeartor $U$, we have the real part (resp. the immaginary part) of $A$ mapped to $real(\lambda)$ resp. $imag(\lambda)$, with $real(A)=(1/2)(A+A^{*})$ and $imag(A)=(1/2i)(A-A^{*})$.

| cite | improve this answer | |
$\endgroup$
  • $\begingroup$ Though I also like to consider normal operators as multiplications, in what way is this answer different from the accepted one? $\endgroup$ – András Bátkai Sep 22 '13 at 18:13
  • $\begingroup$ I think it is more elementary . $\endgroup$ – user36539 Sep 25 '13 at 20:45

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.