decisionSupport (version 1.114)

sample_simple_CPT: Make Conditional Probability tables using the likelihood method

Description

This function creates Conditional Probability Tables for Bayesian Network nodes from parameters that (for complex nodes) can be more easily elicited from experts than the full table. The function uses the Likelihood method. The function combines the make_CPT and sample_CPT functions, but only offers limited flexibility. Refer to the make_CPR and sample_CPT descriptions for details.

Usage

sample_simple_CPT(
  parent_list,
  child_states_n,
  child_prior = NULL,
  b = 2,
  obs_states = NULL
)

Value

list of two data.frames: 1) Conditional Probability Table (CPT); 2) legend table specifying which states of the parent nodes belong to which column in the CPT. If obs_states are given, an additional attribute $sampled specified one random draw, according to the CPT and the obs_states provided.

Arguments

parent_list

named list of parameters for the parent nodes containing a name and a vector of two elements: c(number_of_states,parent_weight).

child_states_n

number of states for the child node.

child_prior

prior distribution for the states of the child node.

b

parameter for the strength of the parent's influence on the child node. A value of 1 causes no response; 3 is quite strong. Defaults to 2.

obs_states

optional vector of observed states for all parents. This has to be complete and names have to correspond exactly with the names of states of the parent nodes. It's also important that the name are given in the exact same sequence as the parents are listed in parent_list.

Author

Eike Luedeling

Examples

parent_list<-list(pare1=c(5,3),parent2=c(3,2),PARE3=c(4,5))
sample_simple_CPT(parent_list,5)
sample_simple_CPT(parent_list,5,obs_states=c("very high","medium","high"))

sample_simple_CPT(parent_list=list(management_intensity=c(5,2),inputs=c(5,1)),5,
     obs_states=c("medium","very high"))$sampled