henhuy
commented
2 years ago Implemented timeseries aggregation.
Aggregation function from aggregate_scalars is reused (and therefore refactored into separate function).
Additionally, a sum_series function has been added to allow summing up series.
Attention: In order to sum series, series column dtype has been changed to np.ndarray - maybe this affects later functions? PLeas check!
Had to make following fixes in test data file:
- index was skipping fifth entry, resulting in different index when comparing dfs
- had to change keyword "ALL" to "All", as aggregation function uses latter version (copied from
aggregate_scalars)
Please review, @jnnr !
Fixes #221.