def test_adjacency_matrix():
g = nx.barbell_graph(8, 5)
cc_ = CreateCliqueComplex(g)
adjacency = nx.adjacency_matrix(g)
assert_almost_equal(adjacency.todense(),
cc_.get_adjacent_matrix().todense())
def test_input_distance_matrix():
dist = distance_matrix(X, X, p=2)
cc_ = CreateCliqueComplex(data=dist, alpha=alpha, data_type='matrix')
cd_ = cc_.create_complex_from_graph()
lap_cc = CreateLaplacianMatrices().fit(cd,
(0, )).transform(cd)[0].todense()
lap_cc_ = CreateLaplacianMatrices().fit(cd_,
(0, )).transform(cd_)[0].todense()
assert_almost_equal(lap_cc, lap_cc_)
def compute_node_edge_entropy(g, i, taus_n, taus_e):
cd = CreateCliqueComplex(graph=g).create_complex_from_graph()
lap = CreateLaplacianMatrices().fit_transform(cd, (0, 1))
n_diff = HeatDiffusion().fit_transform(lap[0], taus_n)
e_diff = HeatDiffusion().fit_transform(lap[1], taus_e)
mh_n = GraphEntropy().fit_transform(n_diff).T
mh_e = GraphEntropy().fit_transform(e_diff).T
if i % 10000 == 0:
print(
"Atoms and Bonds of {} molecules have been embedded...".format(i))
return [mh_n, mh_e]
def test_data_input_with__graph_data_type():
with pytest.raises(ValueError):
CreateCliqueComplex(data=X, alpha=alpha)
def test_graph_input_with_data_type():
with pytest.raises(ValueError):
CreateCliqueComplex(graph=nx.barbell_graph(10, 10), data_type='cloud')
import numpy as np
import pytest
import networkx as nx
from sklearn.exceptions import NotFittedError
from giotto.graphs.create_clique_complex import CreateCliqueComplex,\
CreateLaplacianMatrices, CreateBoundaryMatrices
from numpy.testing import assert_almost_equal
from scipy.spatial import distance_matrix
X = np.random.random((10, 2))
alpha = 0.4
cc = CreateCliqueComplex(data=X, alpha=alpha, data_type='cloud')
cd = cc.create_complex_from_graph()
def test_graph_input_with_data_type():
with pytest.raises(ValueError):
CreateCliqueComplex(graph=nx.barbell_graph(10, 10), data_type='cloud')
def test_data_input_with__graph_data_type():
with pytest.raises(ValueError):
CreateCliqueComplex(data=X, alpha=alpha)
def test_input_distance_matrix():
dist = distance_matrix(X, X, p=2)