tfdf.py_tree.node.NonLeafNode

A non-leaf node i.e.

Inherits From: AbstractNode

tfdf.py_tree.node.NonLeafNode(
    condition: tfdf.py_tree.condition.AbstractCondition,
    pos_child: Optional[tfdf.py_tree.node.AbstractNode] = None,
    neg_child: Optional[tfdf.py_tree.node.AbstractNode] = None,
    value: Optional[tfdf.py_tree.node.AbstractValue] = None
)

a node containing a split/condition.

condition The binary condition of the node.
pos_child The child to visit when the condition is true.
neg_child The child to visit when the condition is false.
value The value/prediction/output of the node if it was a leaf. Not used during prediction.

Methods

pretty

View source

pretty(
    prefix: str, is_pos: Optional[bool], depth: int, max_depth: Optional[int]
) -> str

Returns a recursive readable textual representation of a node.

Args
prefix Prefix printed on the left side. Used to print the surrounding edges.
is_pos True/False if the node is a positive/negative child. None if the node is a root.
depth Depth of the node in the tree. There is no assuption of on the depth of a root.
max_depth Maximum depth for representation. Deeper nodes are skipped.

Returns
A pretty-string representing the node and its children.

scan_structure

View source

scan_structure(
    conditions: tfdf.py_tree.node.ScanStructureAccumulator
)

Scans the structure of the tree.