Arguments

The following arguments are available for the -A options.

t=n: The number of trials. The algorithm repeats the estimation n times, and then returns the best scored structure as the final result.
c= ( 1 | 2 ): Candidate combination.
1 : Parent candidates include only single nodes.
2 (default): Combinations of discrete and continuous nodes are included in parent candidates if both types of nodes are available in the data set.
mins=x₁ [ :x₂ ]: In case of c=2, x₁ and x₂ represent the minimum numbers of candidates for single and combination candidates, respectively. (default: x₁ = x₂ = 0)
skip: Skips the structure search. This is intended to be used with the fixed network for applying parameter estimation under bootstrapped input data.
loops=n: The maximum number of loops of the greedy algorithm. (default: loops=100)
reset= ( on | off ): If on is set, resets the random seed every time before performing bootstrap iteration. (default: on)
bs_check= ( on | off ): If on is set, checks the consistency of the lengths of the secondary IDs and/or the lists for the pid and list mode bootstrap. (default: on)