0.10.0 - 2022-02-04
base.MiniBatchTransformer. Add support for mini-batches to
- Created this module to store estimator unit testing, rather than having it in the
compose.Suffixer that respectively prepend and append a string to the features' name.
compose.Renamer to allow feature renaming following a mapping.
evaluate.progressive_validation to work with
debug_one method to
- Make the
by parameter in
feature_extraction.TargetAgg to be optional, allowing to calculate aggregates over the whole data.
feature_extraction.TargetLagger. Their functionality can be reproduced by combining
feature_extraction.Target now have a
state property. It returns a
pandas.Series representing the current aggregates values within each group.
metrics.ROCAUC works with
- Created this module to store some stuff that was in the
utils module but wasn't necessarily shared between modules.
- Renamed the
Recommender base class into
- Added a
rank method to each recommender.
reco.SurpriseWrapper as it wasn't really useful.
- Added an
is_contextual property to each ranker to indicate if a model makes use of contextual features or not.
stats.Cov each now have an
update_many method which accepts numpy arrays.
utils.Window and use
collections.deque instead where necessary.