Data binning, also called discrete binning or bucketing, is a data pre-processing technique used to reduce the effects of minor observation errors. The original data values which fall into a given small interval, a bin, are replaced by a value representative of that interval, often the central value. It is a form of quantization.
Two classes are implemented here for data binning: BinsEncoder and BinsDiscretizer.
BinsDiscretizeris used to split all values of real feature vector into some bins.BinsEncoderis used to encode each bin with some value.
Example of BinsEncoder usage can be found here.
- Check and add asserts
- Make
encode_binsparameter more flexible may be? - Process empty bins borders correctly
- Process NaNs correctly