BayesianModelInference

class pgmpy.sampling.BayesianModelInference(model)

Bases: Inference

Class to calculate probability (pmf) values specific to Bayesian Models

Parameters:

model: Bayesian Model

model on which inference queries will be computed

pre_compute_reduce(variable)

Get probability arrays for a node as function of conditional dependencies

Internal function used for Bayesian networks, eg. in BayesianModelSampling and BayesianModelProbability.

Parameters:

variable: Bayesian Model Node

node of the Bayesian network

Returns:

dict: dictionary with probability array for node

as function of conditional dependency values

pre_compute_reduce_maps(variable, evidence=None, state_combinations=None)

Get probability array-maps for a node as function of conditional dependencies

Internal function used for Bayesian networks, eg. in BayesianModelSampling and BayesianModelProbability.

Parameters:

variable: Bayesian Model Node

node of the Bayesian network

evidence: list

List of evidence variables to compute the reduced values for. Rest of the parent varaibles of the node are marginalized.

state_combinations: list (default=None)

List of tuple of state combinations for which to compute the reductions maps.

Returns:

dict: dictionary with probability array-index for node as function of conditional dependency values,

dictionary with mapping of probability array-index to probability array.