def test_singly_connected():
"""
Tests generating singly-connected BBN.
:return: None.
"""
g, p = generate_singly_bbn(3)
assert len(g.nodes) == 3
assert len(g.edges) == 2
def test_convert_for_exact_inference():
"""
Tests converting graph and params to a BBN.
:return: None
"""
g, p = generate_singly_bbn(2)
bbn = convert_for_exact_inference(g, p)
assert len(bbn.nodes) == 2
assert len(bbn.edges) == 1
def test_generated_serde():
"""
Tests serde of generated BBN.
:return: Nonde.
"""
g, p = generate_singly_bbn(100, max_iter=10)
e_bbn = convert_for_exact_inference(g, p)
d = Bbn.to_dict(e_bbn)
s = json.dumps(d, sort_keys=True, indent=2)
d = json.loads(s)
o_bbn = Bbn.from_dict(d)
assert len(e_bbn.get_nodes()) == len(o_bbn.get_nodes())
assert len(e_bbn.get_edges()) == len(o_bbn.get_edges())
def test_convert_for_drawing():
"""
Tests converting BBN to networkx DAG.
:return: None
"""
g, p = generate_singly_bbn(2)
bbn = convert_for_exact_inference(g, p)
graph = convert_for_drawing(bbn)
nodes = list(graph.nodes)
edges = list(graph.edges)
assert len(nodes) == 2
assert len(edges) == 1
import numpy as np from pybbn.generator.bbngenerator import generate_singly_bbn, convert_for_exact_inference, convert_for_drawing # very important to set the seed for reproducible results np.random.seed(37) # this method generates the graph, g, and probabilities, p # note we are generating a singly-connected graph g, p = generate_singly_bbn(5, max_iter=5) # you have to convert g and p to a BBN bbn = convert_for_exact_inference(g, p) # you can convert the BBN to a nx graph for visualization nx_graph = convert_for_drawing(bbn)
import json
import numpy as np
from pybbn.generator.bbngenerator import generate_singly_bbn, convert_for_exact_inference
from pybbn.graph.dag import Bbn
np.random.seed(37)
g, p = generate_singly_bbn(900, max_iter=10)
s_bbn = convert_for_exact_inference(g, p)
with open('singly-bbn.json', 'w') as f:
f.write(json.dumps(Bbn.to_dict(s_bbn), sort_keys=True, indent=2))