Open piever opened 6 years ago
Decide which columnwise operations should accept an optional grouping argument. Also clarify what happens for the distributed case (use the groupby implementation for the case by == ()).
groupby
by == ()
Related to this, think how to handle "grouped data".
Similarly, see if it's easy to implement usekey behavior generally when grouping.
usekey
Decide which columnwise operations should accept an optional grouping argument. Also clarify what happens for the distributed case (use the
groupby
implementation for the caseby == ()
).