Returns the JunctionTree computed from the last call to {@link #computeMarginals}. Caller must not modify return value.

Return a random boolean, equally likely to be true or false.

Called after this CommandOption matches an argument. The default implementation simply calls parseArg(String), and returns index+1; unless ind

Samples a subset of elements from this array. @param a @param random @return

Forces this potential to be the identity (all 1s).

Returns the sum of this potential over all cases.

Writes a serialized version of obj to a given file, compressing it using gzip. @param f File to write to @param obj Object to serialize

Compute the L1 residual between two dirichlets

Simple Cholesky decomposition, with no checks on squareness, symmetricality, or positive definiteness. This follows the implementation from JAMA fair

Removes all potentias from this model.

Returns a collection of all the potentials of cliques in the junction tree. (i.e., these are the terms in the numerator of the jounction tre theorem)

This is used in multi-threading. @return A copy of the GEConstraint.

Ensures that <tt>this.inLogSpace == ptl.inLogSpace</tt>. If this is not the case, return a copy of ptl logified or delogified as appropriate. @param

Returns a cluster in the tree that contains var.

Returns the region in this graph whose factor list contains only a given potential. @param ptl @param doCreate If true, an appropriate region will

Return the log(p(parameters)) according to a zero-mean Gaussian with given variance.

Return an list of instances with a particular label.

Get the cost for this node. @return the cost

Returns the degree of a given variable in this factor graph, that is, the number of factors in which the variable is an argument.

Caches some information about this graph that is specific to a given type of inferencer (e.g., a junction tree). @param inferencer Class of inference