Note
Go to the end to download the full example code
Decomposition#
Example of creating a junction tree from a directed graph.
import matplotlib.pyplot as plt
import networkx as nx
B = nx.DiGraph()
B.add_nodes_from(["A", "B", "C", "D", "E", "F"])
B.add_edges_from(
[("A", "B"), ("A", "C"), ("B", "D"), ("B", "F"), ("C", "E"), ("E", "F")]
)
options = {"with_labels": True, "node_color": "white", "edgecolors": "blue"}
fig = plt.figure(figsize=(6, 9))
axgrid = fig.add_gridspec(3, 2)
ax1 = fig.add_subplot(axgrid[0, 0])
ax1.set_title("Bayesian Network")
pos = nx.nx_agraph.graphviz_layout(B, prog="neato")
nx.draw_networkx(B, pos=pos, **options)
mg = nx.moral_graph(B)
ax2 = fig.add_subplot(axgrid[0, 1], sharex=ax1, sharey=ax1)
ax2.set_title("Moralized Graph")
nx.draw_networkx(mg, pos=pos, **options)
jt = nx.junction_tree(B)
ax3 = fig.add_subplot(axgrid[1:, :])
ax3.set_title("Junction Tree")
ax3.margins(0.15, 0.25)
nsize = [2000 * len(n) for n in list(jt.nodes())]
pos = nx.nx_agraph.graphviz_layout(jt, prog="neato")
nx.draw_networkx(jt, pos=pos, node_size=nsize, **options)
plt.tight_layout()
plt.show()
Total running time of the script: ( 0 minutes 0.319 seconds)