This transformer creates an equi-width histogram for each continuous feature column and encodes a discrete category with an automatic bin label for each continuous feature column. The Interval Discretizer is useful when converting continuous features to categorical features so they can be learned by an estimator that supports categorical features natively.
Data Type Compatibility: Continuous
|1||bins||5||int||The number of bins (discrete categories) per continuous feature column.|
Return the possible categories of each feature column:
public categories() : array
Return the intervals of each continuous feature column calculated during fitting:
public intervals() : array
use Rubix\ML\Transformers\IntervalDiscretizer; $transformer = new IntervalDiscretizer(10);