postdist: Calculate mean of posterior parameters and create probability distribution

Description

After finding (learning) posterior parameters, the mean of the distributions are found and used to update the local probability distributions

Usage

postdist(nw) 
postdist.node(nd,nw,vtype = "mode")

Arguments

The network.

The node to update.

vtype

Report "mode" or "mean" of the variance distribution.

Value

A network with updated prob attributes for each node.

Details

In the learning procedure, the postdist.node is called to ensure that the prob-attribute is updated. The prob-attribute is first set in jointprior and is then updated using these functions to update them in the learning procedures. The prob attribute is used by savenet for exporting in .net format.

References

Further information about Deal can be found at: http://www.math.auc.dk/novo/deal.