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
After by_band was implemented in #327, the band_to_calc argument in the analysis function looks redundant. Currently, a user can achieve the same behavior for a few bands with source filtering, and for each band with by_band. In fact, I believe it is worth having the ability to ignore the passband sometimes in order to apply certain types of analysis over all object's sources, as discussed in #325 and #323.
Here, I propose removing band_to_calc from analysis functions and not requiring the band column for them.
The text was updated successfully, but these errors were encountered:
Additional suggestion: we can change batch(by_band: bool) to batch(by_band: bool | BandType | list[BandType]), where BandType is a column dtype for bands, typically str or int. So by_band would also filter bands to have a single band only, batch(by_band='g') or multiple, but not all, bands, batch(by_band=['g', 'r'])
After
by_band
was implemented in #327, theband_to_calc
argument in the analysis function looks redundant. Currently, a user can achieve the same behavior for a few bands with source filtering, and for each band withby_band
. In fact, I believe it is worth having the ability to ignore the passband sometimes in order to apply certain types of analysis over all object's sources, as discussed in #325 and #323.Here, I propose removing
band_to_calc
from analysis functions and not requiring theband
column for them.The text was updated successfully, but these errors were encountered: