Splitter Class#

Splitter #

Splitter(X, Y, criteria)

Splitter class used to create splits of the data.

get_split(indices, feature_indices)

Function that finds the best split of the dataset

Parameters:

Name	Type	Description	Default
`indices`	`memoryview of NDArray`	Indices for which to find a split.	required
`feature_indices`	`memoryview of NDArray`	Features at which to consider splitting.	required

Returns:

Type	Description
`(list, double, int, double, double)`	Returns the best split of the dataset, with the values being: (1) a list containing the left and right indices, (2) the best threshold for doing the splits, (3) what feature to split on, (4) the best criteria score, and (5) the best impurity