Copied from #267:
scikit-hep/boost-histogram#296
suggests that one should be able to pick individual bin indices from a category axis. The suggested command is pick(index1, index2, ...), as suggested by @henryiii
The idea is good, but the implementation cost is not negligible. pick would only work for unordered axes like category. A new kind of reduce constructor must be defined and a corresponding reduce implementation must be written.
Copied from #267:
scikit-hep/boost-histogram#296
suggests that one should be able to pick individual bin indices from a category axis. The suggested command is
pick(index1, index2, ...), as suggested by @henryiiiThe idea is good, but the implementation cost is not negligible.
pickwould only work for unordered axes like category. A new kind of reduce constructor must be defined and a corresponding reduce implementation must be written.