The Node class

These are the collection of different implemented Nodes used by the DecisionTree.

Node

Node(indices, depth, impurity)

The base Node from which all other nodes must inherit.

Parameters:

Name	Type	Description	Default
indices	ndarray	indices in node	required
depth	int	depth of node	required
impurity	float	impurity of node	required

LeafNode

LeafNode(id, indices, depth, impurity, weighted_samples, value, parent)

Parameters:

Name	Type	Description	Default
id	int	unique identifier of leaf node	required
indices	ndarray	indices in leaf node	required
depth	int	depth of leaf node	required
impurity	float	impurity of leaf node	required
weighted_samples	float	summed weight of all samples in leaf node	required
value	ndarray	value of leaf node (depends on LeafBuilder)	required
parent	DecisionNode	parent node	required

DecisionNode

DecisionNode(indices, depth, impurity, threshold, split_idx, left_child=None, right_child=None, parent=None)

Parameters:

Name	Type	Description	Default
indices	ndarray	indices in node	required
depth	int	depth of node	required
impurity	float	impurity in node	required
threshold	float	threshold value of a split	required
split_idx	int	feature index to split on	required
left_child	DecisionNode | LeafNode | None	left child	None
right_child	DecisionNode | LeafNode | None	right child	None
parent	DecisionNode | None	parent node	None