Maxoids.jl
Maxoids.jl provides code to work with conditional independence statements of max-linear Bayeisan networks. This package accompanies the paper Polyhedral aspects of maxoids (arxiv).
Setup
Currently, this package can be installed by executing the following in a Julia REPL.
julia> using Pkg
julia> Pkg.add("https://github.com/ooinaruhugh/PolyhedralAspectsofMaxoids")Quick start
julia> using Maxoids
julia> G = complete_DAG(4)
Directed graph with 4 nodes and the following edges:
(1, 2)(1, 3)(1, 4)(2, 3)(2, 4)(3, 4)
julia> maxoid(G, [1,2,3,4,5,6])
7-element Vector{CIStmt}:
[1 _||_ 3 | 2]
[1 _||_ 3 | {2, 4}]
[1 _||_ 4 | 2]
[1 _||_ 4 | 3]
[1 _||_ 4 | {2, 3}]
[2 _||_ 4 | 3]
[2 _||_ 4 | {1, 3}]