Structure Scoring

Structure score classes used by score-based discovery workflows.

Base Class and Utilities

BaseStructureScore(data[, state_names])

Abstract base class for structure scoring in pgmpy.

get_scoring_method(scoring_method, data)

Discrete Scores

K2(data[, state_names])	K2 structure score for discrete Bayesian networks using uniform Dirichlet priors.
BDeu(data[, equivalent_sample_size, state_names])	BDeu structure score for discrete Bayesian networks with Dirichlet priors.
BDs(data[, equivalent_sample_size, state_names])	BDs structure score for discrete Bayesian networks.
LogLikelihood(data[, state_names])	Log-likelihood structure score for discrete Bayesian networks.
AIC(data[, state_names])	AIC structure score for discrete Bayesian networks.
BIC(data[, state_names])	BIC structure score for discrete Bayesian networks.

Gaussian Scores

LogLikelihoodGauss(data[, state_names])	Log-likelihood structure score for Gaussian Bayesian networks.
AICGauss(data[, state_names])	AIC structure score for Gaussian Bayesian networks.
BICGauss(data[, state_names])	BIC structure score for Gaussian Bayesian networks.

Conditional-Gaussian Scores

LogLikelihoodCondGauss(data[, state_names])	Log-likelihood score for Bayesian networks with mixed discrete and continuous variables.
AICCondGauss(data[, state_names])	AIC structure score for Bayesian networks with mixed discrete and continuous variables.
BICCondGauss(data[, state_names])	BIC structure score for Bayesian networks with mixed discrete and continuous variables.