Old engine for Continuous Time Bayesian Networks. Superseded by reCTBN. 🐍
https://github.com/madlabunimib/PyCTBN
You can not select more than 25 topics
Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
178 lines
5.5 KiB
178 lines
5.5 KiB
|
|
# License: MIT License
|
|
|
|
|
|
import glob
|
|
import math
|
|
import os
|
|
import unittest
|
|
|
|
import networkx as nx
|
|
import numpy as np
|
|
import psutil
|
|
|
|
import json
|
|
import pandas as pd
|
|
|
|
|
|
from pyctbn.legacy.structure_graph.sample_path import SamplePath
|
|
from pyctbn.legacy.estimators.structure_constraint_based_estimator import StructureConstraintBasedEstimator
|
|
from pyctbn.legacy.utility.sample_importer import SampleImporter
|
|
|
|
import copy
|
|
|
|
|
|
class TestStructureConstraintBasedEstimator(unittest.TestCase):
|
|
@classmethod
|
|
def setUpClass(cls):
|
|
with open("./tests/data/networks_and_trajectories_binary_data_01_3.json") as f:
|
|
raw_data = json.load(f)
|
|
|
|
trajectory_list_raw= raw_data[0]["samples"]
|
|
|
|
trajectory_list = [pd.DataFrame(sample) for sample in trajectory_list_raw]
|
|
|
|
variables= pd.DataFrame(raw_data[0]["variables"])
|
|
prior_net_structure = pd.DataFrame(raw_data[0]["dyn.str"])
|
|
|
|
|
|
cls.importer = SampleImporter(
|
|
trajectory_list=trajectory_list,
|
|
variables=variables,
|
|
prior_net_structure=prior_net_structure
|
|
)
|
|
|
|
cls.importer.import_data()
|
|
#cls.s1 = sp.SamplePath(cls.importer)
|
|
|
|
#cls.traj = cls.s1.concatenated_samples
|
|
|
|
# print(len(cls.traj))
|
|
cls.s1 = SamplePath(cls.importer)
|
|
cls.s1.build_trajectories()
|
|
cls.s1.build_structure()
|
|
|
|
def test_structure_1(self):
|
|
true_edges = copy.deepcopy(self.s1.structure.edges)
|
|
true_edges = set(map(tuple, true_edges))
|
|
|
|
se1 = StructureConstraintBasedEstimator(self.s1,0.1,0.1)
|
|
edges = se1.estimate_structure(processes_number=2)
|
|
|
|
self.assertFalse(se1.spurious_edges())
|
|
self.assertEqual(edges, true_edges)
|
|
|
|
|
|
def test_structure_2(self):
|
|
with open("./tests/data/networks_and_trajectories_binary_data_02_10_1.json") as f:
|
|
raw_data = json.load(f)
|
|
|
|
trajectory_list_raw= raw_data["samples"]
|
|
|
|
trajectory_list = [pd.DataFrame(sample) for sample in trajectory_list_raw]
|
|
|
|
variables= pd.DataFrame(raw_data["variables"])
|
|
prior_net_structure = pd.DataFrame(raw_data["dyn.str"])
|
|
|
|
|
|
self.importer = SampleImporter(
|
|
trajectory_list=trajectory_list,
|
|
variables=variables,
|
|
prior_net_structure=prior_net_structure
|
|
)
|
|
|
|
self.importer.import_data()
|
|
#cls.s1 = sp.SamplePath(cls.importer)
|
|
|
|
#cls.traj = cls.s1.concatenated_samples
|
|
|
|
# print(len(cls.traj))
|
|
self.s1 = SamplePath(self.importer)
|
|
self.s1.build_trajectories()
|
|
self.s1.build_structure()
|
|
|
|
true_edges = copy.deepcopy(self.s1.structure.edges)
|
|
true_edges = set(map(tuple, true_edges))
|
|
|
|
se1 = StructureConstraintBasedEstimator(self.s1,0.1,0.1)
|
|
edges = se1.estimate_structure(True)
|
|
|
|
|
|
'calculate precision and recall'
|
|
n_missing_edges = 0
|
|
n_added_fake_edges = 0
|
|
|
|
|
|
n_added_fake_edges = len(edges.difference(true_edges))
|
|
|
|
n_missing_edges = len(true_edges.difference(edges))
|
|
|
|
n_true_positive = len(true_edges) - n_missing_edges
|
|
|
|
precision = n_true_positive / (n_true_positive + n_added_fake_edges)
|
|
|
|
recall = n_true_positive / (n_true_positive + n_missing_edges)
|
|
|
|
self.assertGreaterEqual(precision,0.75)
|
|
self.assertGreaterEqual(recall,0.75)
|
|
|
|
|
|
def test_structure_3(self):
|
|
with open("./tests/data/networks_and_trajectories_ternary_data_01_6_1.json") as f:
|
|
raw_data = json.load(f)
|
|
|
|
trajectory_list_raw= raw_data["samples"]
|
|
|
|
trajectory_list = [pd.DataFrame(sample) for sample in trajectory_list_raw]
|
|
|
|
variables= pd.DataFrame(raw_data["variables"])
|
|
prior_net_structure = pd.DataFrame(raw_data["dyn.str"])
|
|
|
|
|
|
self.importer = SampleImporter(
|
|
trajectory_list=trajectory_list,
|
|
variables=variables,
|
|
prior_net_structure=prior_net_structure
|
|
)
|
|
|
|
self.importer.import_data()
|
|
#cls.s1 = sp.SamplePath(cls.importer)
|
|
|
|
#cls.traj = cls.s1.concatenated_samples
|
|
|
|
# print(len(cls.traj))
|
|
self.s1 = SamplePath(self.importer)
|
|
self.s1.build_trajectories()
|
|
self.s1.build_structure()
|
|
|
|
true_edges = copy.deepcopy(self.s1.structure.edges)
|
|
true_edges = set(map(tuple, true_edges))
|
|
|
|
se1 = StructureConstraintBasedEstimator(self.s1,0.1,0.1)
|
|
edges = se1.estimate_structure(True)
|
|
|
|
|
|
'calculate precision and recall'
|
|
n_missing_edges = 0
|
|
n_added_fake_edges = 0
|
|
|
|
|
|
n_added_fake_edges = len(edges.difference(true_edges))
|
|
|
|
n_missing_edges = len(true_edges.difference(edges))
|
|
|
|
n_true_positive = len(true_edges) - n_missing_edges
|
|
|
|
precision = n_true_positive / (n_true_positive + n_added_fake_edges)
|
|
|
|
recall = n_true_positive / (n_true_positive + n_missing_edges)
|
|
|
|
self.assertGreaterEqual(precision,0.75)
|
|
self.assertGreaterEqual(recall,0.75)
|
|
|
|
|
|
|
|
|
|
|
|
if __name__ == '__main__':
|
|
unittest.main()
|
|
|