Open stress-tess opened 4 months ago
I'm not sure how useful this would be in general... I might just implement the chapel side logic for my purpose and punt on making it user facing for now. In particular I want to be able to pass different number of elements to select from each group and specifying the order of that k-array seems a bit hard. The natural way would be according to the order from g.unique_keys
I'll hold off on actually making this a user facing function until I can discuss with the rest of the team. If we did go this route, I imagine it would be most similar to unique
or sample
should we add
mink
andmaxk
aggregations to groupby, which return a segarray. Most of the logic needed for this was implemented forfind
to return all occurrences in issue https://github.com/Bears-R-Us/arkouda/issues/3009