I have just started my first proper course in DSP, and am a little confused about a statement in my textbook. The book states:
"Notice that when the Nyqust-frequency is defined as $\pi/\Delta t$, we will with $\omega_m = 2 \pi m/N \Delta t$ where $m = 0, \pm 1, \pm2, ...$ have $\omega_m \neq \pi/ \Delta t$, unless $N$ is even. If we are working with symmetric functions, it can be wise to use an odd number of sampling points."
I see the logic in the first sentence here, but don't understand how it follows from this that it is wise to use an odd number of sampling points for a symmetric function. This statement is not elaborated upon further. Can anyone explain this logic to me? Why it is wise to use an odd number of sampling points for a symmetric function? If anyone could explain this, I would greatly appreciate it!