This question is duplicate of the question Find projection-valued measure associated with parity operator.\

But in that question @Jacky Chong does not state how he found the operator \begin{align} P_\pi(\lambda) =P_{\pi}((-\infty,\lambda]) =\delta(\lambda-1)P_\text{even}+\delta(\lambda+1)P_\text{odd} \end{align}

Also we know that $P(R)=Id \enspace$ or we should have $P(\lambda_1)\leq P(\lambda_2)$ for $\lambda_1 < \lambda _2$ but if we put 2 and 1 into the $P_{\pi}$ we get $0:L^2\rightarrow L^2$ and $P_{even}$ resp.

My question is could someone explain me how to find $P_{\pi}$ explicitly using maybe Stieltjes inversion formula or something else

(1) I need 50 reps to comment on the problem i linked
(2) I already found the spectrum and resolvent of parity operator.They're $$\sigma(\Pi)=\{1,-1\} \enspace R_{\Pi}(z)=\frac{\Pi+zI}{1-z^2}$$ (3) This is Problem 3.2 from Teschl G. - Mathematical methods in quantum mechanics


This is simply the standard way of writing a finite sum as an integral relative to a corresponding atomic measure. In particular: if the operator in question can be written as a countable sum $$ A = \sum_i \lambda_i P_i $$ (with the $P_i$ equal to spectral projectors and hence mutually orthogonal and "complete"), then the measure associated with $A$ can be written as $$ P_A(\lambda) = \sum_i \delta(\lambda - \lambda_i)P_i. $$ To see that this works, it suffices to show that for a Borel function $f$, $f(A)$ is indeed equal to $\int_{\Bbb R} f(\lambda) \,dP_{A}(\lambda)$. In other words, this is simply a consequence of the fact that $$ f\left(\sum_i \lambda_i P_i\right) = \sum_i f(\lambda_i) P_i. $$

| cite | improve this answer | |
  • $\begingroup$ First thank you for your answer.So by sayig "spectral projectors" what you mean is their sum is identity like if all the $\lambda_i=1$ and by making some of the $\lambda_i=0$ or something else we get $A$.In our case we have $P_{even}$ and $P_{odd}$ and their $\lambda$'s 1 and -1 resp. Also $\delta$ is dirac-delta right?It is not heaviside or whatever the name it is.Could you explain why it suffices to show that $f(A)=\int_{\mathbb{R}}f(\lambda)dP_A(\lambda)$ because it forms a $C*$ homomorphism ? but there could be some other things that might be $C*$ homo without being PVM. $\endgroup$ – Oğuzhan Kılıç Oct 28 at 8:45
  • $\begingroup$ and i wrote an example that if we had $\lambda_2=2$ we get $0$ map but we should have had a map that extends $P_{even}$.Maybe i'm understanding it wrong but im confused little bit :D.Also sorry for late return $\endgroup$ – Oğuzhan Kılıç Oct 28 at 8:45
  • $\begingroup$ For your first comment: all your statements are correct. The precise answer to why it suffices depends on what exactly it means for a projection valued measure to be "associated with" an operator. I'm rusty with all this and don't have a reference on hand; it would be helpful if you could consult your textbook or notes for how exactly this is defined. $\endgroup$ – Ben Grossmann Oct 29 at 1:22
  • $\begingroup$ I'm not exactly sure what you're saying about your example $\endgroup$ – Ben Grossmann Oct 29 at 1:23
  • $\begingroup$ i kind of get it now. I was confusing identity function and 1 function.Thanks for your answers anyway $\endgroup$ – Oğuzhan Kılıç Oct 29 at 9:08

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.