Expand source code

from typing import Generator

from ..tree import Tree


class Initializer:
    """
    The abstract interface for the tree initializers.

    Initializers are responsible for setting the starting values of the model, in particular:
      - structure of decision and leaf nodes
      - variables and values used in splits
      - values of leaf nodes

    Good initialization of trees helps speed up convergence of sampling

    Default behaviour is to leave trees uninitialized
    """

    def __init__(self):
        self.n_trees = 1

    def initialize_tree(self, tree: Tree, tree_number: int) -> None:
        pass

    def initialize_trees(self, trees: Generator[Tree, None, None]) -> None:
        n_trees = 0
        for tree_number, tree in enumerate(trees):
            self.initialize_tree(tree, tree_number)
            n_trees += 1
        self.n_trees = n_trees

Classes

class Initializer

The abstract interface for the tree initializers.

Initializers are responsible for setting the starting values of the model, in particular: - structure of decision and leaf nodes - variables and values used in splits - values of leaf nodes

Good initialization of trees helps speed up convergence of sampling

Default behaviour is to leave trees uninitialized

Expand source code

class Initializer:
    """
    The abstract interface for the tree initializers.

    Initializers are responsible for setting the starting values of the model, in particular:
      - structure of decision and leaf nodes
      - variables and values used in splits
      - values of leaf nodes

    Good initialization of trees helps speed up convergence of sampling

    Default behaviour is to leave trees uninitialized
    """

    def __init__(self):
        self.n_trees = 1

    def initialize_tree(self, tree: Tree, tree_number: int) -> None:
        pass

    def initialize_trees(self, trees: Generator[Tree, None, None]) -> None:
        n_trees = 0
        for tree_number, tree in enumerate(trees):
            self.initialize_tree(tree, tree_number)
            n_trees += 1
        self.n_trees = n_trees

Subclasses

SklearnTreeInitializer

Methods

def initialize_tree(self, tree: Tree, tree_number: int) ‑> None

Expand source code

def initialize_tree(self, tree: Tree, tree_number: int) -> None:
    pass

def initialize_trees(self, trees: Generator[Tree, None, None]) ‑> None

Expand source code

def initialize_trees(self, trees: Generator[Tree, None, None]) -> None:
    n_trees = 0
    for tree_number, tree in enumerate(trees):
        self.initialize_tree(tree, tree_number)
        n_trees += 1
    self.n_trees = n_trees