Interpreting 'catnet' package's cnProb output for a catnetwork object

Question

I am new to Bayesian Networks. I am trying to use catnet package in R, however, I am having difficulty understanding the output of cnProb() function. For instance, here is a new catnetwork object:

cnet_test <- cnNew(
  nodes = c("a", "b", "c"),
  cats = list(c("1","2"), c("1","2"), c("1","2")),
  parents = list(NULL, c(1), c(1,2)))

This should then result in a network like this, right? I am assuming parents = parameter here means node X is parent of ...

However, when doing cnProb() on this catnet object, it returns the following:

$a
   1    2 
0.19 0.81 
$b
  a 1     2    
A 1 0.396 0.604
B 2 0.611 0.389
$c
  a b 1     2    
A 1 1 0.519 0.481
B 1 2 0.878 0.122
A 2 1 0.666 0.334
B 2 2 0.89  0.11 

This seems the exact opposite of the network diagram. According to the documentation, cnprob:

Returns the list of conditional probabilities of nodes specified by which parameter of a catNetwork object. Node probabilities are reported in the following format. First, node name and its parents are given, then a list of probability values corresponding to all combination of parent categories (put in brackets) and node categories. For example, the conditional probability of a node with two parents, such that both the node and its parents have three categories, is given by 27 values, one for each of the 333 combination.

I am wondering how exactly do we interpret the output from cnProb? or is my interpretation of cnNew's parents parameter wrong. Any information will be helpful.

Answer 1

Your interpretation of "parents" parameter in cnNew() is incorrect and your diagram does not corresponds to the network you actually define. "parents = list(NULL, c(1), c(1,2))" means that "1" is parent of "2" and "1 and 2" are parents of "3". So, the network is {a->b; a->c, b->c}.