koho::Tree Class Reference

Binary tree structure build up of nodes. More...

#include <decision_tree.h>

Collaboration diagram for koho::Tree:

Public Member Functions
	Tree (OutputsIdx_t n_outputs, std::vector< ClassesIdx_t > n_classes, FeaturesIdx_t n_features)
	Create a new tree without nodes. More...
	Tree (OutputsIdx_t n_outputs, ClassesIdx_t *n_classes_ptr, FeaturesIdx_t n_features)
	Create a new tree without nodes for Python binding. More...
	Tree ()
	Create a new tree without nodes for Python binding using pickle. More...
NodesIdx_t	add_node (TreeDepthIdx_t depth, NodesIdx_t parent_id, bool is_left, FeaturesIdx_t feature, int NA, Features_t threshold, const std::vector< std::vector< Histogram_t >> &histogram, double impurity, double improvement)
	Add a new node to the tree. More...
void	predict (Features_t *X, SamplesIdx_t n_samples, double *y_prob)
	Predict classes probabilities for the test data. More...
void	calculate_feature_importances (double *importances)
	Calculate feature importances from the decision tree. More...
void	serialize (std::ofstream &fout)
	Serialize. More...
void	deserialize (std::ifstream &fin)
	Deserialize. More...

Public Attributes
OutputsIdx_t	n_outputs
std::vector< ClassesIdx_t >	n_classes
ClassesIdx_t	n_classes_max
FeaturesIdx_t	n_features
TreeDepthIdx_t	max_depth
NodesIdx_t	node_count
std::vector< Node >	nodes

Detailed Description

Binary tree structure build up of nodes.

Constructor & Destructor Documentation

Tree() [1/3]

koho::Tree::Tree	(	OutputsIdx_t	n_outputs,
		std::vector< ClassesIdx_t >	n_classes,
		FeaturesIdx_t	n_features
	)

Create a new tree without nodes.

Parameters

[in]	n_outputs	Number of outputs.
[in]	n_classes	Number of classes for each output.
[in]	n_features	Number of features.

Tree() [2/3]

koho::Tree::Tree	(	OutputsIdx_t	n_outputs,
		ClassesIdx_t *	n_classes_ptr,
		FeaturesIdx_t	n_features
	)

Create a new tree without nodes for Python binding.

Parameters

[in]	n_outputs	Number of outputs.
[in]	n_classes	Number of classes for each output.
[in]	n_features	Number of features.

Tree() [3/3]

koho::Tree::Tree ( )

inline

Create a new tree without nodes for Python binding using pickle.

Member Function Documentation

add_node()

NodesIdx_t koho::Tree::add_node	(	TreeDepthIdx_t	depth,
		NodesIdx_t	parent_id,
		bool	is_left,
		FeaturesIdx_t	feature,
		int	NA,
		Features_t	threshold,
		const std::vector< std::vector< Histogram_t >> &	histogram,
		double	impurity,
		double	improvement
	)

Add a new node to the tree.

The new node registers itself as the child of its parent.

calculate_feature_importances()

void koho::Tree::calculate_feature_importances

(

double *

importances

)

Calculate feature importances from the decision tree.

Parameters

[in,out]

importances

Feature importances corresponding to all features [n_features].

deserialize()

void koho::Tree::deserialize

(

std::ifstream &

fin

)

Deserialize.

predict()

void koho::Tree::predict	(	Features_t *	X,
		SamplesIdx_t	n_samples,
		double *	y_prob
	)

Predict classes probabilities for the test data.

Parameters

[in]	X	Test input samples [n_samples x n_features].
[in]	n_samples	Number of samples in the test data.
[in,out]	y_prob	Class probabilities corresponding to the test input samples [n_samples x n_outputs x n_classes_max]. We use n_classes_max to create a nice 3D array to hold the predicted values x samples x classes as the number of classes can be different for different outputs.

Using 1d array addressing for X and y_prob to support efficient Cython bindings to Python using memory views.

serialize()

void koho::Tree::serialize

(

std::ofstream &

fout

)

Serialize.

Member Data Documentation

max_depth

TreeDepthIdx_t koho::Tree::max_depth

n_classes

std::vector<ClassesIdx_t> koho::Tree::n_classes

n_classes_max

ClassesIdx_t koho::Tree::n_classes_max

n_features

FeaturesIdx_t koho::Tree::n_features

n_outputs

OutputsIdx_t koho::Tree::n_outputs

node_count

NodesIdx_t koho::Tree::node_count

nodes

std::vector<Node> koho::Tree::nodes

---

The documentation for this class was generated from the following files:

• decision_tree.h
• decision_tree.cpp