Test case is more expressive now

test/test_jpt/test_jpt.py

@@ -7,6 +7,7 @@
 import networkx as nx
 import numpy as np
 import pandas as pd
+import portion
 import random_events
 import sklearn.datasets
 from jpt import infer_from_dataframe as old_infer_from_dataframe
@@ -389,6 +390,7 @@ def setUpClass(cls):
 
         model_sl_sw = JPT(variables, min_samples_leaf=0.4, features=[cls.sl, cls.sw], targets=variables)
         model_sl_sw.fit(df)
+
         cls.model_sl_sw = model_sl_sw.marginal([cls.sl, cls.sw])
 
         model_pl_pw = JPT(variables, min_samples_leaf=0.4, features=[cls.pl, cls.pw], targets=variables)
@@ -444,7 +446,12 @@ def test_to_bayesian_network(self):
         root = DiscreteDistribution(self.species_latent_variable, self.model_species.weights)
         self.assertEqual(root.weights, [1 / 3] * 3)
         bayesian_network.add_node(root)
-        self.assertEqual(bayesian_network.probability(Event()), 1.)
+
+        p_species = ConditionalProbabilisticCircuit(self.model_species.variables)
+        p_species.from_unit(self.model_species)
+        bayesian_network.add_node(p_species)
+        bayesian_network.add_edge(root, p_species)
+
 
         # mount the interaction term with the latent variable of the sepal distribution
         p_sepal_species = ConditionalProbabilityTable(self.sepal_latent_variable)
@@ -479,7 +486,16 @@ def test_to_bayesian_network(self):
         self.assertEqual(bayesian_network.probability(Event()), 1.)
         self.assertAlmostEqual(bayesian_network.as_probabilistic_circuit().probability(Event()), 1)
 
-        p_specie_1 = Event({self.species_latent_variable: 0})
-        self.assertAlmostEqual(bayesian_network.probability(p_specie_1), 1 / 3)
-        self.assertAlmostEqual(bayesian_network.as_probabilistic_circuit().probability(p_specie_1), 1 / 3)
+        e_species_1 = Event({self.species: 0})
+        bn_p_species_1 = bayesian_network.probability(e_species_1)
+        self.assertAlmostEqual(bn_p_species_1, 1 / 3)
+        self.assertAlmostEqual(bayesian_network.as_probabilistic_circuit().probability(e_species_1), 1 / 3)
+
+        complex_event = Event({self.species: 0,
+                               self.sl: portion.closed(4.5, 5.5)})
+        pc = bayesian_network.as_probabilistic_circuit()
+        pc_m = pc.marginal([v for v in pc.variables if not v.name.endswith(".latent")]).simplify()
+        self.assertEqual(pc_m.variables, (self.pl, self.pw, self.sl, self.sw, self.species))
 
+        self.assertAlmostEqual(pc_m.probability(complex_event), 0.2333333)
+        self.assertLess(len(pc_m.weighted_edges), math.prod([len(v.domain) for v in bayesian_network.variables]))