Describes the main functions for scoring a node within a Gaussian Network using the bge scoring metric. More...

#include <stdio.h>
#include <stdlib.h>
#include <math.h>
#include <string.h>
#include "bge_score.h"
#include "bge_matrix.h"

Include dependency graph for bge_score.c:

Functions
double *	create_log_gamma_ratio_table (int samples, int alpha_omega, int no_vars)

double	log_prefactor (int no_samples, int alpha_mu)

double	LogBgeScore (int node, unsigned int family, int no_parents, int alpha_mu, int alpha_omega, double log_prefactor, double log_gamma_ratio_table, Bge_Matrix prior_matrix, Bge_Matrix posterior_matrix, Bge_Matrix *data)
	Computes The log of the bge score for a node either with or without parents. More...

double	LogBgeScoreWithoutParents (int node, int alpha_mu, int alpha_omega, double log_prefactor, double log_gamma_ratio_table, Bge_Matrix prior_matrix, Bge_Matrix posterior_matrix, Bge_Matrix data)
	Computes The log of the bge score for a node without parents. More...

double	LogBgeScoreWithParents (int node, unsigned int family, int no_parents, int alpha_mu, int alpha_omega, double log_prefactor, double log_gamma_ratio_table, Bge_Matrix prior_matrix, Bge_Matrix posterior_matrix, Bge_Matrix *data)
	Computes The log of the bge score for a node with parents. More...

Detailed Description

Describes the main functions for scoring a node within a Gaussian Network using the bge scoring metric.

Function Documentation

Computes The log of the bge score for a node either with or without parents.

Parameters

node	The integer representation of the node in relation to the row/column in the dataset
family	The family of the node being score where the value family[0] = node
no_parents	The number of parents of the node
alpha_mu	A hyper parameter used to in computing the posterior_matrix
alpha_omega	A hyper parameter used to in computing the posterior_matrix
log_prefactor	The ratio of logarithms for the score
log_gamma_ratio_table	A table that contains all the possible gamma_ratios for each different size parent set
prior_matrix	The prior matrix T
posterior_matrix	The posterior matrix R
data	- The continuous data that the network is being learned from

double LogBgeScoreWithoutParents	(	int	node,
		int	alpha_mu,
		int	alpha_omega,
		double	log_prefactor,
		double *	log_gamma_ratio_table,
		Bge_Matrix *	prior_matrix,
		Bge_Matrix *	posterior_matrix,
		Bge_Matrix *	data
	)

Computes The log of the bge score for a node without parents.

Parameters

node	The integer representation of the node in relation to the row/column in the dataset
alpha_mu	A hyper parameter used to in computing the posterior_matrix
alpha_omega	A hyper parameter used to in computing the posterior_matrix
log_prefactor	The ratio of logarithms for the score
log_gamma_ratio_table	A table that contains all the possible gamma_ratios for each different size parent set
prior_matrix	The prior matrix T
posterior_matrix	The posterior matrix R
data	- The continuous data that the network is being learned from

double LogBgeScoreWithParents	(	int	node,
		unsigned int *	family,
		int	no_parents,
		int	alpha_mu,
		int	alpha_omega,
		double	log_prefactor,
		double *	log_gamma_ratio_table,
		Bge_Matrix *	prior_matrix,
		Bge_Matrix *	posterior_matrix,
		Bge_Matrix *	data
	)

Computes The log of the bge score for a node with parents.

Parameters

node	The integer representation of the node in relation to the row/column in the dataset
family	The family of the node being score where the value family[0] = node
no_parents	The number of parents of the node
alpha_mu	A hyper parameter used to in computing the posterior_matrix
alpha_omega	A hyper parameter used to in computing the posterior_matrix
log_prefactor	The ratio of logarithms for the score
log_gamma_ratio_table	A table that contains all the possible gamma_ratios for each different size parent set
prior_matrix	The prior matrix T
posterior_matrix	The posterior matrix R
data	- The continuous data that the network is being learned from