You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Is there anything against deprecating the BeamformerTimeSq class and all other *Sq classes in acoular.tbeamform?
The result functions have already been merged anyway to a single function in BeamformerTime and BeamformerTimeTraj and we currently have 4 extra classes implementing the sq logic.
Instead, we could introduce a squared_output / squared_result or simply sq attribute, which is False by default. r_diag would then only be able to be used when sq=True.
The only thing I am afraid of is that people think they apply "diagonal removal" but don't set the squared attribute to True. We could avoid this by changing the default value of r_diag to False and by printing a warning if r_diag==True and sq==False.
The text was updated successfully, but these errors were encountered:
I think this change is unnecessary. The implementation of the *Sq classes is relatively lightweight and removing them has not much effect except breaking backwards compatibility.
@esarradj, @gherold
Is there anything against deprecating the
BeamformerTimeSq
class and all other*Sq
classes inacoular.tbeamform
?The
result
functions have already been merged anyway to a single function inBeamformerTime
andBeamformerTimeTraj
and we currently have 4 extra classes implementing the sq logic.Instead, we could introduce a
squared_output
/squared_result
or simplysq
attribute, which isFalse
by default.r_diag
would then only be able to be used whensq=True
.The only thing I am afraid of is that people think they apply "diagonal removal" but don't set the squared attribute to True. We could avoid this by changing the default value of
r_diag
toFalse
and by printing a warning ifr_diag==True
andsq==False
.The text was updated successfully, but these errors were encountered: