Fitting a slicer to a grouped dataset, say grouped by mode, currently makes it fail on features that have not been grouped by mode.
In a typical case, SliceIntervals obtained from some grouped Feature to be used for slicing would have the levels corpus, piece, mode, intervals. In principle, it will be possible to use Feature.apply_slice_intervals() even when a feature does not have a mode level (i.e., it hasn't been grouped by mode, e.g. because it's Notes that don't come with keys by default).
Suggested solution
Dropping the mode level from the SliceIntervals, applying it to the feature, re-adding the mode level through a join operation.
Fitting a slicer to a grouped dataset, say grouped by mode, currently makes it fail on features that have not been grouped by mode.
In a typical case,
SliceIntervalsobtained from some grouped Feature to be used for slicing would have the levelscorpus, piece, mode, intervals. In principle, it will be possible to useFeature.apply_slice_intervals()even when a feature does not have amodelevel (i.e., it hasn't been grouped by mode, e.g. because it'sNotesthat don't come with keys by default).Suggested solution
Dropping the
modelevel from the SliceIntervals, applying it to the feature, re-adding the mode level through a join operation.