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
Currently, we can batch sub-light curves with batch(by_band=True). It is a great feature, but I believe we would benefit from some refactoring of it.
Me feeling is that current batch interface is over-loaded and has too many combinations of 1) analysis function type, 2) its Dask meta types, 3) grouping-by flavors and 4) types of the output.
I propose to introduce batch_by(by=column_name) and its thin wrapper batch_by_band to replace the current batch(by_band=True) interface. It would simplify our interfaces, make the code more maintainable, and allow users to use any categorical column, not band only.
The text was updated successfully, but these errors were encountered:
Currently, we can batch sub-light curves with
batch(by_band=True)
. It is a great feature, but I believe we would benefit from some refactoring of it.Me feeling is that current
batch
interface is over-loaded and has too many combinations of 1) analysis function type, 2) its Dask meta types, 3) grouping-by flavors and 4) types of the output.I propose to introduce
batch_by(by=column_name)
and its thin wrapperbatch_by_band
to replace the currentbatch(by_band=True)
interface. It would simplify our interfaces, make the code more maintainable, and allow users to use any categorical column, not band only.The text was updated successfully, but these errors were encountered: