col_gen_estimator.Node

class col_gen_estimator.Node[source]

All node related information.

Attributes:
id: (int) ID of the node.
candidate_splits: (list(int)) IDs of the candidate splits that can be used

at this node.

candidate_splits_count: (list(int)) Count of how many times the

corresponding split was seen in initialization. This is useful to trim the candidate_splits list. Used only at the initialization and not during training.

last_split: (int) ID of the last split added to the candidate_splits. Used

only at the initialization and not during training.

parent: (int) ID of the parent node. Parent of the root node is -1.
left_child: (int) ID of the left child node. -1 if the child is a leaf.
right_child: (int) ID of the right child node. -1 if the child is a leaf.
children_are_leaves: (bool) True if the children of this node are leaves.
__init__() → None[source]

Examples using col_gen_estimator.Node

sphx_glr_auto_examples_dtree_experiment.py

Code to launch decision tree experiments.