def test_vertices_in_e_not_v(self):
e = pd.DataFrame([['ING', 'ABN'],
['BNP', 'ABN'],
['RAB', 'ABN'],
['ABN', 'BNP'],
['BNP', 'ING']],
columns=['creditor', 'debtor'])
with pytest.raises(Exception) as e_info:
ge.Graph(self.v, e, v_id='name', src='creditor', dst='debtor')
msg = 'Some source vertices are not in v.'
assert e_info.value.args[0] == msg
e = pd.DataFrame([['ING', 'ABN'],
['BNP', 'ABN'],
['ING', 'RAB'],
['ABN', 'BNP'],
['BNP', 'ING']],
columns=['creditor', 'debtor'])
with pytest.raises(Exception) as e_info:
ge.Graph(self.v, e, v_id='name', src='creditor', dst='debtor')
msg = 'Some destination vertices are not in v.'
assert e_info.value.args[0] == msg
def test_vertices_with_no_edge(self):
v = pd.DataFrame([['ING'], ['ABN'], ['BNP'], ['RAB']],
columns=['name'])
with pytest.warns(UserWarning, match='RAB vertex has no edges.'):
ge.Graph(v, self.e, v_id='name', src='creditor', dst='debtor')
v = pd.DataFrame([['ING'], ['ABN'], ['BNP'], ['RAB'], ['UBS']],
columns=['name'])
with pytest.warns(UserWarning, match=r' vertices have no edges.'):
ge.Graph(v, self.e, v_id='name', src='creditor', dst='debtor')
def test_init_class(self):
g = ge.Graph(self.v_s, self.e_s, v_id='name', src='creditor',
dst='debtor', edge_label='type')
assert isinstance(g, ge.sGraph)
assert isinstance(g, ge.LabelGraph)
assert g.num_edges == 4
def test_strength_by_label(self):
g = ge.Graph(self.v, self.e, v_id=['name', 'country'],
src=['creditor', 'c_country'],
dst=['debtor', 'd_country'],
edge_label=['type', 'EUR'],
weight='value')
s = np.rec.array([(0, 0, 1e6),
(0, 1, 1e4 + 1e6 + 3e3),
(0, 2, 1e4),
(0, 3, 3e3),
(1, 1, 2.3e7),
(1, 2, 2.3e7),
(2, 1, 7e5),
(2, 3, 7e5),
(3, 0, 4e5),
(3, 1, 4e5)],
dtype=[('label', np.uint8),
('id', np.uint8),
('value', np.float64)])
s_test = g.strength_by_label(get=True)
assert np.all(s_test == s), s_test
assert np.all(g.lv.strength == s), g.v.strength
def test_to_networkx_orig(self):
g = ge.Graph(self.v, self.e, v_id=['name', 'country'],
src=['creditor', 'c_country'],
dst=['debtor', 'd_country'],
edge_label=['type', 'EUR'],
weight='value', v_group='country')
v = [(('ING', 'NL'), {'group': 'NL'}),
(('ABN', 'NL'), {'group': 'NL'}),
(('BNP', 'FR'), {'group': 'FR'}),
(('BNP', 'IT'), {'group': 'IT'})]
e = [(('ING', 'NL'), ('ABN', 'NL'),
{'weight': 1e6, 'label': ('interbank', False)}),
(('ABN', 'NL'), ('BNP', 'FR'),
{'weight': 1e4, 'label': ('interbank', False)}),
(('ABN', 'NL'), ('ING', 'NL'),
{'weight': 4e5, 'label': ('external', True)}),
(('BNP', 'FR'), ('ABN', 'NL'),
{'weight': 2.3e7, 'label': ('external', False)}),
(('BNP', 'IT'), ('ABN', 'NL'),
{'weight': 3e3, 'label': ('interbank', False)}),
(('BNP', 'IT'), ('ABN', 'NL'),
{'weight': 7e5, 'label': ('interbank', True)})]
gx = g.to_networkx(original=True)
assert np.all(list(gx.nodes(data=True)) == v)
assert np.all(list(gx.edges(data=True)) == e)
def test_num_edges_by_label(self):
g = ge.Graph(self.v, self.e, v_id=['name', 'country'],
src=['creditor', 'c_country'],
dst=['debtor', 'd_country'],
edge_label=['type', 'EUR'])
test_num = np.array([3, 1, 1, 1], dtype='u1')
assert np.all(g.num_edges_label == test_num)
def test_init(self):
g = ge.Graph(self.v, self.e, v_id='name', src='creditor',
dst='debtor', weight='value')
assert isinstance(g, ge.sGraph)
assert isinstance(g, ge.WeightedGraph)
assert np.all(g.e == self._e), g.e
def test_to_sparse(self):
g = ge.Graph(self.v, self.e, v_id=['name', 'country'],
src=['creditor', 'c_country'],
dst=['debtor', 'd_country'],
edge_label=['type', 'EUR'])
mat = [np.array([[0, 1, 0, 0],
[0, 0, 1, 0],
[0, 0, 0, 0],
[0, 1, 0, 0]]),
np.array([[0, 0, 0, 0],
[0, 0, 0, 0],
[0, 1, 0, 0],
[0, 0, 0, 0]]),
np.array([[0, 0, 0, 0],
[0, 0, 0, 0],
[0, 0, 0, 0],
[0, 1, 0, 0]]),
np.array([[0, 0, 0, 0],
[1, 0, 0, 0],
[0, 0, 0, 0],
[0, 0, 0, 0]])]
for i in range(len(mat)):
assert np.all(g.adjacency_matrix()[i].toarray() == mat[i])
def test_multi_id_init(self):
v = pd.DataFrame([['ING', 'NL'],
['ABN', 'NL'],
['BNP', 'FR'],
['BNP', 'IT']],
columns=['name', 'country'])
e = pd.DataFrame([['ING', 'NL', 'ABN', 'NL'],
['BNP', 'FR', 'ABN', 'NL'],
['ABN', 'NL', 'BNP', 'IT'],
['BNP', 'IT', 'ING', 'NL']],
columns=['creditor', 'c_country',
'debtor', 'd_country'])
g = ge.Graph(v, e, v_id=['name', 'country'],
src=['creditor', 'c_country'],
dst=['debtor', 'd_country'])
test_e = np.sort(np.rec.array([(0, 1), (2, 1), (1, 3), (3, 0)],
dtype=[('src', np.uint8), ('dst', np.uint8)]))
test_dict = {('ING', 'NL'): 0, ('ABN', 'NL'): 1,
('BNP', 'FR'): 2, ('BNP', 'IT'): 3}
assert np.all(g.e == test_e), g.e
assert test_dict == g.id_dict
def test_in_strength(self):
g = ge.Graph(self.v, self.e, v_id='name', src='creditor',
dst='debtor', weight='value')
s_in = np.array([0, 1e6 + 1.7e5, 1e4])
s_test = g.in_strength(get=True)
assert np.all(s_test == s_in), s_test
assert np.all(g.v.in_strength == s_in), g.v.in_strength
def test_out_strength(self):
g = ge.Graph(self.v, self.e, v_id='name', src='creditor',
dst='debtor', weight='value')
s_out = np.array([1e6, 1e4, 1.7e5])
s_test = g.out_strength(get=True)
assert np.all(s_test == s_out), s_test
assert np.all(g.v.out_strength == s_out), g.v.out_strength
def test_to_sparse(self):
g = ge.Graph(self.v, self.e, v_id='name', src='creditor',
dst='debtor', weight='value')
mat = np.array([[0, 1e6, 0],
[0, 0, 1e4],
[0, 1.7e5, 0]])
assert np.all(g.adjacency_matrix().toarray() == mat)
def test_out_degree(self):
g = ge.Graph(self.v, self.e, v_id='name',
src='creditor', dst='debtor')
d_out = np.array([1, 1, 2])
d_test = g.out_degree(get=True)
assert np.all(d_test == d_out), d_test
assert np.all(g.v.out_degree == d_out), g.v.out_degree
def test_in_degree(self):
g = ge.Graph(self.v, self.e, v_id='name',
src='creditor', dst='debtor')
d_in = np.array([1, 2, 1])
d_test = g.in_degree(get=True)
assert np.all(d_test == d_in), d_test
assert np.all(g.v.in_degree == d_in), g.v.in_degree
def test_total_weight(self):
g = ge.Graph(self.v, self.e, v_id=['name', 'country'],
src=['creditor', 'c_country'],
dst=['debtor', 'd_country'],
edge_label=['type', 'EUR'],
weight='value')
assert g.total_weight == 1e6 + 2.3e7 + 7e5 + 3e3 + 1e4 + 4e5
def test_degree_init(self):
v = pd.DataFrame([['ING'], ['ABN'], ['BNP'], ['RAB'], ['UBS']],
columns=['name'])
d = np.array([2, 2, 2, 0, 0])
with pytest.warns(UserWarning):
g = ge.Graph(v, self.e, v_id='name', src='creditor', dst='debtor')
assert np.all(g.v.degree == d), g.v.degree
def test_init(self):
g = ge.Graph(self.v, self.e, v_id='name',
src='creditor', dst='debtor')
test_e = np.sort(np.rec.array([(0, 1), (2, 1), (1, 2), (2, 0)],
dtype=[('src', np.uint8), ('dst', np.uint8)]))
assert isinstance(g, ge.sGraph)
assert isinstance(g, ge.DirectedGraph)
assert np.all(g.e == test_e), g.e
def test_vertex_group(self):
g = ge.Graph(self.v, self.e, v_id='name', src='creditor',
dst='debtor', weight='value', v_group='country')
test_v = np.array([0, 0, 1], dtype=np.uint8)
assert isinstance(g, ge.sGraph)
assert isinstance(g, ge.WeightedGraph)
assert np.all(g.v.group == test_v), g.v.group
def test_duplicated_vertices(self):
v = pd.DataFrame([['ING'], ['ABN'], ['BNP'], ['ABN']],
columns=['name'])
with pytest.raises(Exception) as e_info:
ge.Graph(v, self.e, v_id='name', src='creditor', dst='debtor')
msg = 'There is at least one repeated id in the vertex dataframe.'
assert e_info.value.args[0] == msg
def test_total_weight_by_label(self):
g = ge.Graph(self.v, self.e, v_id=['name', 'country'],
src=['creditor', 'c_country'],
dst=['debtor', 'd_country'],
edge_label=['type', 'EUR'],
weight='value')
w_by_l = np.array([1e6 + 3e3 + 1e4, 2.3e7, 7e5, 4e5], dtype='f8')
test = g.total_weight_label
assert np.all(test == w_by_l), test
def test_in_degree(self):
g = ge.Graph(self.v, self.e, v_id=['name', 'country'],
src=['creditor', 'c_country'],
dst=['debtor', 'd_country'],
edge_label=['type', 'EUR'])
d_in = np.array([1, 3, 1, 0])
d_test = g.in_degree(get=True)
assert np.all(d_test == d_in), d_test
assert np.all(g.v.in_degree == d_in), g.v.in_degree
def test_in_strength(self):
g = ge.Graph(self.v, self.e, v_id=['name', 'country'],
src=['creditor', 'c_country'],
dst=['debtor', 'd_country'],
edge_label=['type', 'EUR'],
weight='value')
s_in = np.array([4e5, 1e6 + 2.3e7 + 7e5 + 3e3, 1e4, 0])
s_test = g.in_strength(get=True)
assert np.all(s_test == s_in), s_test
assert np.all(g.v.in_strength == s_in), g.v.in_strength
def test_out_strength(self):
g = ge.Graph(self.v, self.e, v_id=['name', 'country'],
src=['creditor', 'c_country'],
dst=['debtor', 'd_country'],
edge_label=['type', 'EUR'],
weight='value')
s_out = np.array([1e6, 1e4 + 4e5, 2.3e7, 7e5 + 3e3])
s_test = g.out_strength(get=True)
assert np.all(s_test == s_out), s_test
assert np.all(g.v.out_strength == s_out), g.v.out_strength
def test_to_sparse_compressed(self):
g = ge.Graph(self.v, self.e, v_id=['name', 'country'],
src=['creditor', 'c_country'],
dst=['debtor', 'd_country'],
edge_label=['type', 'EUR'])
mat = np.array([[0, 1, 0, 0],
[1, 0, 1, 0],
[0, 1, 0, 0],
[0, 1, 0, 0]], dtype=float)
assert np.all(g.adjacency_matrix(compressed=True).toarray() == mat)
def test_to_sparse_compressed(self):
g = ge.Graph(self.v, self.e, v_id=['name', 'country'],
src=['creditor', 'c_country'],
dst=['debtor', 'd_country'],
edge_label=['type', 'EUR'],
weight='value')
mat = np.array([[0, 1e6, 0, 0],
[4e5, 0, 1e4, 0],
[0, 2.3e7, 0, 0],
[0, 7e5 + 3e3, 0, 0]])
assert np.all(g.adjacency_matrix(compressed=True).toarray() == mat)
def test_vertex_group(self):
v = pd.DataFrame([['ING', 'NL'], ['ABN', 'NL'], ['BNP', 'FR']],
columns=['name', 'country'])
g = ge.Graph(v, self.e, v_id='name',
src='creditor', dst='debtor', v_group='country')
test_v = np.array([0, 0, 1], dtype=np.uint8)
assert isinstance(g, ge.sGraph)
assert isinstance(g, ge.DirectedGraph)
assert np.all(g.v.group == test_v), g.v.group
def test_to_networkx(self):
g = ge.Graph(self.v, self.e, v_id='name',
src='creditor', dst='debtor')
v = [0, 1, 2]
e = [(0, 1), (1, 2), (2, 0), (2, 1)]
adj = {0: {1: {}}, 1: {2: {}}, 2: {0: {}, 1: {}}}
gx = g.to_networkx()
assert np.all(list(gx.nodes) == v)
assert np.all(list(gx.edges) == e)
assert gx.adj == adj
def test_vertex_group(self):
g = ge.Graph(self.v, self.e, v_id=['name', 'country'],
src=['creditor', 'c_country'],
dst=['debtor', 'd_country'],
edge_label=['type', 'EUR'],
v_group='country')
test_v = np.array([0, 0, 1, 2], dtype=np.uint8)
assert isinstance(g, ge.sGraph)
assert isinstance(g, ge.LabelGraph)
assert np.all(g.v.group == test_v), g.v.group
def test_init_edges(self):
g = ge.Graph(self.v_s, self.e_s, v_id='name', src='creditor',
dst='debtor', edge_label='type')
test_e = np.sort(np.rec.array([(0, 0, 1),
(1, 2, 1),
(0, 2, 1),
(0, 1, 2),
(1, 1, 0)],
dtype=[('label', np.uint8),
('src', np.uint8),
('dst', np.uint8)]))
assert np.all(g.e == test_e), g.e
def test_duplicated_edges(self):
e = pd.DataFrame([['ING', 'ABN'],
['BNP', 'ABN'],
['ING', 'ABN'],
['ABN', 'BNP'],
['BNP', 'ING']],
columns=['creditor', 'debtor'])
with pytest.raises(Exception) as e_info:
ge.Graph(self.v, e, v_id='name', src='creditor', dst='debtor')
msg = 'There are repeated edges.'
assert e_info.value.args[0] == msg