Inference in Markov random fields

Inference is the process that consists in computing new probabilistc information from a Markov random field and some evidence. aGrUM/pyAgrum mainly focus and the computation of (joint) posterior for some variables of the Markov random fields given soft or hard evidence that are the form of likelihoods on some variables. Inference is a hard task (NP-complete). For now, aGrUM/pyAgrum implements only one exact inference for Markov random field.

Shafer-Shenoy Inference in Markov random field

class pyAgrum.ShaferShenoyMRFInference(MN, use_binary_join_tree=True)

Class used for Shafer-Shenoy inferences for Markov random field.

ShaferShenoyMRFInference(bn) -> ShaferShenoyMRFInference
Parameters:

  • mrf (pyAgrum.MarkovRandomField) – a Markov random field

Parameters:

  • MN (IMarkovRandomField)

  • use_binary_join_tree (bool)

H(X)

Deprecated I in LazyPropagation/ShaferShenoyMRFInference

I(X, Y)

Deprecated I in LazyPropagation/ShaferShenoyMRFInference

MN()

Deprecated method in gum.ShaferShenoyMRFInference for pyAgrum>1.5.2

MRF()
VI(X, Y)

Deprecated VI in LazyPropagation/ShaferShenoyMRFInference

addAllTargets()

Add all the nodes as targets.

Return type:

None

addEvidence(*args)

Adds a new evidence on a node (might be soft or hard).

Parameters:

  • id (int) – a node Id

  • nodeName (int) – a node name

  • val – (int) a node value

  • val – (str) the label of the node value

  • vals (list) – a list of values

Raises:
Return type:

None

addJointTarget(targets)

Add a list of nodes as a new joint target. As a collateral effect, every node is added as a marginal target.

Parameters:

  • list – a list of names of nodes

  • targets (object)

Raises:

pyAgrum.UndefinedElement – If some node(s) do not belong to the Bayesian network

Return type:

None

addTarget(*args)

Add a marginal target to the list of targets.

Parameters:

  • target (int) – a node Id

  • nodeName (str) – a node name

Raises:

pyAgrum.UndefinedElement – If target is not a NodeId in the Bayes net

Return type:

None

chgEvidence(*args)

Change the value of an already existing evidence on a node (might be soft or hard).

Parameters:

  • id (int) – a node Id

  • nodeName (int) – a node name

  • val (intstr) – a node value or the label of the node value

  • vals (List[float]) – a list of values

Raises:
Return type:

None

eraseAllEvidence()

Removes all the evidence entered into the network.

Return type:

None

eraseAllJointTargets()

Clear all previously defined joint targets.

Return type:

None

eraseAllMarginalTargets()

Clear all the previously defined marginal targets.

Return type:

None

eraseAllTargets()

Clear all previously defined targets (marginal and joint targets).

As a result, no posterior can be computed (since we can only compute the posteriors of the marginal or joint targets that have been added by the user).

Return type:

None

eraseEvidence(*args)

Remove the evidence, if any, corresponding to the node Id or name.

Parameters:

  • id (int) – a node Id

  • nodeName (int) – a node name

Raises:

pyAgrum.IndexError – If the node does not belong to the Bayesian network

Return type:

None

eraseJointTarget(targets)

Remove, if existing, the joint target.

Parameters:

  • list – a list of names or Ids of nodes

  • targets (object)

Raises:

  • pyAgrum.IndexError – If one of the node does not belong to the Bayesian network

  • pyAgrum.UndefinedElement – If node Id is not in the Bayesian network

Return type:

None

eraseTarget(*args)

Remove, if existing, the marginal target.

Parameters:

  • target (int) – a node Id

  • nodeName (int) – a node name

Raises:

  • pyAgrum.IndexError – If one of the node does not belong to the Bayesian network

  • pyAgrum.UndefinedElement – If node Id is not in the Bayesian network

Return type:

None

evidenceImpact(target, evs)

Create a pyAgrum.Potential for P(target|evs) (for all instanciation of target and evs)

Parameters:

  • target (set) – a set of targets ids or names.

  • evs (set) – a set of nodes ids or names.

Warning

if some evs are d-separated, they are not included in the Potential.

Returns:

a Potential for P(targets|evs)

Return type:

pyAgrum.Potential

evidenceJointImpact(*args)

Create a pyAgrum.Potential for P(joint targets|evs) (for all instanciation of targets and evs)

Parameters:

  • targets (List[intstr]) – a list of node Ids or node names

  • evs (Set[intstr]) – a set of nodes ids or names.

Returns:

a Potential for P(target|evs)

Return type:

pyAgrum.Potential

Raises:

pyAgrum.Exception – If some evidene entered into the Bayes net are incompatible (their joint proba = 0)

evidenceProbability()
Returns:

the probability of evidence

Return type:

float

getNumberOfThreads()

returns the number of threads used by LazyPropagation during inferences.

Returns:

the number of threads used by LazyPropagation during inferences

Return type:

int

hardEvidenceNodes()
Returns:

the set of nodes with hard evidence

Return type:

set

hasEvidence(*args)
Parameters:

  • id (int) – a node Id

  • nodeName (str) – a node name

Returns:

True if some node(s) (or the one in parameters) have received evidence

Return type:

bool

Raises:

pyAgrum.IndexError – If the node does not belong to the Bayesian network

hasHardEvidence(nodeName)
Parameters:

  • id (int) – a node Id

  • nodeName (str) – a node name

Returns:

True if node has received a hard evidence

Return type:

bool

Raises:

pyAgrum.IndexError – If the node does not belong to the Bayesian network

hasSoftEvidence(*args)
Parameters:

  • id (int) – a node Id

  • nodeName (str) – a node name

Returns:

True if node has received a soft evidence

Return type:

bool

Raises:

pyAgrum.IndexError – If the node does not belong to the Bayesian network

isGumNumberOfThreadsOverriden()

Indicates whether LazyPropagation currently overrides aGrUM’s default number of threads (see method setNumberOfThreads).

Returns:

A Boolean indicating whether LazyPropagation currently overrides aGrUM’s default number of threads

Return type:

bool

isJointTarget(targets)
Parameters:

  • list – a list of nodes ids or names.

  • targets (object)

Returns:

True if target is a joint target.

Return type:

bool

Raises:

  • pyAgrum.IndexError – If the node does not belong to the Bayesian network

  • pyAgrum.UndefinedElement – If node Id is not in the Bayesian network

isTarget(*args)
Parameters:

  • variable (int) – a node Id

  • nodeName (str) – a node name

Returns:

True if variable is a (marginal) target

Return type:

bool

Raises:

  • pyAgrum.IndexError – If the node does not belong to the Bayesian network

  • pyAgrum.UndefinedElement – If node Id is not in the Bayesian network

joinTree()
Returns:

the current join tree used

Return type:

pyAgrum.CliqueGraph

jointMutualInformation(targets)
Parameters:

targets (object)

Return type:

float

jointPosterior(targets)

Compute the joint posterior of a set of nodes.

Parameters:

list – the list of nodes whose posterior joint probability is wanted

Warning

The order of the variables given by the list here or when the jointTarget is declared can not be assumed to be used by the Potential.

Returns:

a const ref to the posterior joint probability of the set of nodes.

Return type:

pyAgrum.Potential

Raises:

pyAgrum.UndefinedElement – If an element of nodes is not in targets

Parameters:

targets (object)

jointTargets()
Returns:

the list of target sets

Return type:

list

junctionTree()
Returns:

the current junction tree

Return type:

pyAgrum.CliqueGraph

makeInference()

Perform the heavy computations needed to compute the targets’ posteriors

In a Junction tree propagation scheme, for instance, the heavy computations are those of the messages sent in the JT. This is precisely what makeInference should compute. Later, the computations of the posteriors can be done ‘lightly’ by multiplying and projecting those messages.

Return type:

None

nbrEvidence()
Returns:

the number of evidence entered into the Bayesian network

Return type:

int

nbrHardEvidence()
Returns:

the number of hard evidence entered into the Bayesian network

Return type:

int

nbrJointTargets()
Returns:

the number of joint targets

Return type:

int

nbrSoftEvidence()
Returns:

the number of soft evidence entered into the Bayesian network

Return type:

int

nbrTargets()
Returns:

the number of marginal targets

Return type:

int

posterior(*args)

Computes and returns the posterior of a node.

Parameters:

  • var (int) – the node Id of the node for which we need a posterior probability

  • nodeName (str) – the node name of the node for which we need a posterior probability

Returns:

a const ref to the posterior probability of the node

Return type:

pyAgrum.Potential

Raises:

pyAgrum.UndefinedElement – If an element of nodes is not in targets

setEvidence(evidces)

Erase all the evidences and apply addEvidence(key,value) for every pairs in evidces.

Parameters:

evidces (dict) – a dict of evidences

Raises:
setMaxMemory(gigabytes)

sets an upper bound on the memory consumption admissible

Parameters:

gigabytes (float) – this upper bound in gigabytes.

Return type:

None

setNumberOfThreads(nb)

If the argument nb is different from 0, this number of threads will be used during inferences, hence overriding aGrUM’s default number of threads. If, on the contrary, nb is equal to 0, the parallelized inference engine will comply with aGrUM’s default number of threads.

Parameters:

nb (int) – the number of threads to be used by ShaferShenoyMRFInference

Return type:

None

setTargets(targets)

Remove all the targets and add the ones in parameter.

Parameters:

targets (set) – a set of targets

Raises:

pyAgrum.UndefinedElement – If one target is not in the Bayes net

softEvidenceNodes()
Returns:

the set of nodes with soft evidence

Return type:

set

targets()
Returns:

the list of marginal targets

Return type:

list

property thisown

The membership flag

updateEvidence(evidces)

Apply chgEvidence(key,value) for every pairs in evidces (or addEvidence).

Parameters:

evidces (dict) – a dict of evidences

Raises: