pgm_learner¶

Message types¶

• pgm_learner/DiscreteNodeState
• pgm_learner/LinearGaussianGraphState
• pgm_learner/GraphEdge
• pgm_learner/DiscreteNode
• pgm_learner/LinearGaussianNode
• pgm_learner/LinearGaussianNodeState
• pgm_learner/ConditionalProbability
• pgm_learner/GraphStructure
• pgm_learner/DiscreteGraphState

pgm_learner/DiscreteNodeState¶

Field:	node (string) – state (string) –

string node
string state

pgm_learner/LinearGaussianGraphState¶

Field:	node_states[] (pgm_learner/LinearGaussianNodeState) –

pgm_learner/LinearGaussianNodeState[] node_states

pgm_learner/GraphEdge¶

Field:	node_from (string) – node_to (string) –

string node_from
string node_to

pgm_learner/DiscreteNode¶

Field:	name (string) – outcomes[] (string) – parents[] (string) – children[] (string) – CPT[] (pgm_learner/ConditionalProbability) –

string name
string[] outcomes
string[] parents
string[] children
pgm_learner/ConditionalProbability[] CPT

pgm_learner/LinearGaussianNode¶

Field:	name (string) – parents[] (string) – children[] (string) – mean (float32) – variance (float32) – mean_scalar[] (float32) –

string name
string[] parents
string[] children

float32 mean
float32 variance

float32[] mean_scalar # scalar for parents

pgm_learner/LinearGaussianNodeState¶

Field:	node (string) – state (float32) –

string node
float32 state

pgm_learner/ConditionalProbability¶

Field:	values[] (string) – probabilities[] (float32) –

string[]  values
float32[] probabilities

pgm_learner/GraphStructure¶

Field:	nodes[] (string) – edges[] (pgm_learner/GraphEdge) –

string[] nodes
pgm_learner/GraphEdge[] edges

pgm_learner/DiscreteGraphState¶

Field:	node_states[] (pgm_learner/DiscreteNodeState) –

pgm_learner/DiscreteNodeState[] node_states

Service types¶

• pgm_learner/DiscreteQuery
• pgm_learner/LinearGaussianParameterEstimation
• pgm_learner/LinearGaussianStructureEstimation
• pgm_learner/DiscreteStructureEstimation
• pgm_learner/DiscreteParameterEstimation

pgm_learner/DiscreteQuery¶

Field (Request):
	nodes[] (pgm_learner/DiscreteNode) – evidence[] (pgm_learner/DiscreteNodeState) – query[] (string) –
Field (Response):
	nodes[] (pgm_learner/DiscreteNode) –

pgm_learner/DiscreteNode[] nodes # information of each nodes
pgm_learner/DiscreteNodeState[] evidence # evidnce
string[] query # query
---
pgm_learner/DiscreteNode[] nodes # information of each nodes

pgm_learner/LinearGaussianParameterEstimation¶

Field (Request):
	graph (pgm_learner/GraphStructure) – states[] (pgm_learner/LinearGaussianGraphState) –
Field (Response):
	nodes[] (pgm_learner/LinearGaussianNode) –

pgm_learner/GraphStructure graph # graph skeleton
pgm_learner/LinearGaussianGraphState[] states # trial data
---
pgm_learner/LinearGaussianNode[] nodes # information of each nodes

pgm_learner/LinearGaussianStructureEstimation¶

Field (Request):
	states[] (pgm_learner/LinearGaussianGraphState) – pvalparam (float64) – bins (uint16) – indegree (uint16) –
Field (Response):
	graph (pgm_learner/GraphStructure) –

pgm_learner/LinearGaussianGraphState[] states # trial data
float64 pvalparam # optional
uint16 bins # optional
uint16 indegree # optional
---
pgm_learner/GraphStructure graph # structure of network

pgm_learner/DiscreteStructureEstimation¶

Field (Request):
	states[] (pgm_learner/DiscreteGraphState) – pvalparam (float64) – indegree (uint16) –
Field (Response):
	graph (pgm_learner/GraphStructure) –

pgm_learner/DiscreteGraphState[] states # trial data
float64 pvalparam # optional
uint16 indegree # optional
---
pgm_learner/GraphStructure graph # structure of network

pgm_learner/DiscreteParameterEstimation¶

Field (Request):
	graph (pgm_learner/GraphStructure) – states[] (pgm_learner/DiscreteGraphState) –
Field (Response):
	nodes[] (pgm_learner/DiscreteNode) –

pgm_learner/GraphStructure graph # graph skeleton
pgm_learner/DiscreteGraphState[] states # trial data
---
pgm_learner/DiscreteNode[] nodes # information of each nodes