cnNodeLoglik: Node Log-likelihood
In sdnet: Soft-Discretization-Based Bayesian Network Inference

Description Usage Arguments Value Author(s) See Also Examples

For a given data sample, the function calculates the log-likelihood of a node with respect to a specified parent set.

1	cnNodeLoglik(object, node, data, pert=NULL, softmode=FALSE, klmode=FALSE)

`object`	a `catNetwork` object
`node`	an `integer` or a `list` of `integer`s, node indices in the `data`
`data`	a `matrix` or `data.frame` of categories
`pert`	an optional perturbation `matrix` or `data.frame`
`softmode`	a `logical`, turns on/off the soft quantization mode
`klmode`	a `logical`, use the negative KL-distance or the log-likelihood

a numeric value

N. Balov

cnLoglik

1
2
3

  cnet <- cnRandomCatnet(numnodes=10, maxpars=3, numcats=2)
  psamples <- cnSamples(object=cnet, numsamples=100)
  cnNodeLoglik(cnet, node=5, data=psamples)