<< Click to Display Table of Contents >> Navigation:  Using SMILE Wrappers > Dynamic Bayesian networks > Temporal evidence

To specify evidence for the plate nodes in the DBN, use Network.setTemporalEvidence and setTemporalVirtualEvidence. The code snippet below sets the temporal evidence in slices 5 and 7. Node represented by evidenceNodeHandle is assumed to have two outcomes.

Java:

net.setTemporalEvidence(evidenceNodeHandle, 5, 1); 

net.setTemporalVirtualEvidence(evidenceNodeHandle, 7,
    new double[] { 0.4, 0.6 });

Python:

net.set_temporal_evidence(evidenceNodeHandle, 5, 1)

net.set_temporal_virtual_evidence(evidenceNodeHandle, 7, [0.4,0.6])

R:

net$setTemporalEvidence(evidenceNodeHandle, 5, 1)

net$setTemporalVirtualEvidence(evidenceNodeHandle, 7,
    c(0.4, 0.6))

C#:

net.SetTemporalEvidence(evidenceNodeHandle, 5, 1); 

net.SetTemporalVirtualEvidence(evidenceNodeHandle, 7,

    new double[] { 0.4, 0.6 });

To retrieve the temporal evidence, use Network.getTemporalEvidence and Network.getTemporalVirtualEvidence.

Other useful methods are Network.hasTemporalEvidence and isTemporalEvidence, which check whether a node has any temporal evidence or temporal evidence in specified temporal order, respectively.