implemented spectral response correction and calculate spectral response correction

[Lulus792]

Implemented into CommonLumi the spectral_response_correction aswell as calculate_reponse_correction. Using a helper function _apply which takes an operator besides the ref signal since both function only differ in that functionality. Added tests aswell.

Description of the change

• implemented spectral_response_correction
• implemented calculate_spectral_response

Progress of the PR

• Change implemented (can be split into several points),
• docstring updated (if appropriate),
• added tests,
• add a changelog entry in the upcoming_changes folder (see upcoming_changes/README.rst),
• ready for review.

Minimal example of the bug fix or the new feature

import lumispy as lum
import numpy as np
s = lum.signals.LumiSpectrum(np.arange(10))
ref = lum.signals.LumiSpectrum(0.2 * np.ones_like(np.arange(10)))
s.spectral_response_correction(ref, inplace=True)

[ericpre]

How different is it from the following:

import lumispy as lum
import numpy as np

s = lum.signals.LumiSpectrum(np.arange(10))
ref = lum.signals.LumiSpectrum(0.2 * np.ones_like(np.arange(10)))
s = s * ref

If there is no difference, and if the https://peps.python.org/pep-0020/ is followed, s = s * ref should be preferred because it is simple and explicit and the method spectral_response_correction is not obvious what correction it would be applying.

[Lulus792]

from what i know the idea was to implement two functions one for multiplication and the other for division. where for spectral_response_correction you want to remove the reference signal that you get from the measurement instrument which is apperently done via multiplication. And the same exist with division tho i am also not quite sure. Thats at least what i understould tho i would have to ask @jlaehne

[Lulus792]

i suppose there isnt a difference its i suppose simpöy to be more meanigful? Tho as insaid not entirely sure tho it should be the same as your example

[jlaehne]

How different is it from the following:

import lumispy as lum
import numpy as np

s = lum.signals.LumiSpectrum(np.arange(10))
ref = lum.signals.LumiSpectrum(0.2 * np.ones_like(np.arange(10)))
s = s * ref

If there is no difference, and if the https://peps.python.org/pep-0020/ is followed, s = s * ref should be preferred because it is simple and explicit and the method spectral_response_correction is not obvious what correction it would be applying.

Indeed, the function should use s = s * ref in the end, but will become a wrapper to handle cases of mismatching signal axes using interpolate_on_axis. ref should be cast onto the axis of s. And if s has a larger range than ref, it's axis should be restricted to the range where the mathematical operation can be applied.