Fix cross spectrum plotting

[matteobachetti]

Eliminates the annoying "imaginary" curve in Powerspectrum and other derivates of Crossspectrum that don't have one. Also, fixes the labels

[codecov]

Codecov Report

Merging #763 (23ad8a5) into main (0f746f6) will increase coverage by 5.40%.
The diff coverage is 100.00%.

@@            Coverage Diff             @@
##             main     #763      +/-   ##
==========================================
+ Coverage   90.74%   96.14%   +5.40%     
==========================================
  Files          43       43              
  Lines        7918     7937      +19     
==========================================
+ Hits         7185     7631     +446     
+ Misses        733      306     -427     

Files	Coverage Δ
stingray/crossspectrum.py	99.04% <100.00%> (+0.03%)	⬆️

... and 14 files with indirect coverage changes

📣 We’re building smart automated test selection to slash your CI/CD build times. Learn more

[mgullik]

Hi @matteobachetti,

this is useful.
One small thing, real and imaginary parts can have very different values, in that case one of the two quantities could be very difficult to read in the plot.
Have you thought of using
ax2 = ax.twinx()
ax2.plot(self.freq, self.power.imag, marker, color="g", alpha=0.5, label="Imaginary Part")
ax2.tick_params('y', colors="g")

and then

if labels is not None:
ax2.set_ylabel(labels[1])
else:
ax2.set_ylabel(f"Power ({self.norm})")

Or something on this line?

[matteobachetti]

[matteobachetti]

@mgullik the failures are unrelated to this PR