Welcome to sklearn-features’s documentation!

sklearn-features provides an API to simplify feature engineering with scikit-learn and pandas.

API summary

transformers.DataFrameSelector(key) Transforms a DataFrame into a Series by selecting a single column by key.
transformers.SeriesReshaper Transforms a Series of size N into an (N, 1) shaped numpy array.
transformers.DataFrameReshaper Transforms a DataFrame of size NxM into an (N, M) shaped numpy array.
transformers.NullTransformer([feature_name]) Pass through the Series completely unchanged.
transformers.ScalingTransformer(scaling_factor) Apply a constant scaling factor to a Series.
transformers.DateAttributeTransformer(attr) Select a particular attribute from the .dt property of a Series.
transformers.MultiDateTransformer(dates)
transformers.LinearDateTransformer([d0, delta]) Convert a datetime Series into a float Series.
transformers.LabelEncoderWithUnknown([...]) Convert a categorical feature into values [0, n], where [0, n) represent the known categories from the training data and n represents unknown data.
transformers.OneHotWithUnknown([feature_names])
transformers.OneHotWithFixedFeatures([...])
transformers.series_pipeline(key, steps)
transformers.dataframe_pipeline(key, steps)

Indices and tables