def test_APPNP_linkmodel_apply_sparse():
G, features = create_graph_features()
adj = G.to_adjacency_matrix()
features, adj = GCN_Aadj_feats_op(features, adj)
adj = adj.tocoo()
A_indices = np.expand_dims(np.hstack((adj.row[:, None], adj.col[:, None])), 0)
A_values = np.expand_dims(adj.data, 0)
generator = FullBatchLinkGenerator(G, sparse=True, method="gcn")
appnpnModel = APPNP(
layer_sizes=[3], activations=["relu"], generator=generator, dropout=0.5
)
x_in, x_out = appnpnModel.in_out_tensors()
model = keras.Model(inputs=x_in, outputs=x_out)
# Check fit method
out_indices = np.array([[[0, 1], [1, 2]]], dtype="int32")
preds_1 = model.predict([features[None, :, :], out_indices, A_indices, A_values])
assert preds_1.shape == (1, 2, 2, 3)
# Check fit method
preds_2 = model.predict(generator.flow([("a", "b"), ("b", "c")]))
assert preds_2.shape == (1, 2, 2, 3)
assert preds_1 == pytest.approx(preds_2)
def test_APPNP_linkmodel_apply_dense():
G, features = create_graph_features()
adj = nx.to_numpy_array(G)[None, :, :]
n_nodes = features.shape[0]
nodes = G.nodes()
node_features = pd.DataFrame.from_dict(
{n: f
for n, f in zip(nodes, features)}, orient="index")
G = StellarGraph(G, node_features=node_features)
generator = FullBatchLinkGenerator(G, sparse=False, method="none")
appnpnModel = APPNP([3], generator, activations=["relu"], dropout=0.5)
x_in, x_out = appnpnModel.build()
model = keras.Model(inputs=x_in, outputs=x_out)
# Check fit method
out_indices = np.array([[[0, 1], [1, 2]]], dtype="int32")
preds_1 = model.predict([features[None, :, :], out_indices, adj])
assert preds_1.shape == (1, 2, 2, 3)
# Check fit_generator method
preds_2 = model.predict_generator(generator.flow([("a", "b"), ("b", "c")]))
assert preds_2.shape == (1, 2, 2, 3)
assert preds_1 == pytest.approx(preds_2)
def test_generator_flow_targets_as_list(self):
generator = FullBatchLinkGenerator(self.G)
link_ids = list(self.G.edges())[:3]
link_targets = [1] * len(link_ids)
gen = generator.flow(link_ids, link_targets)
inputs, y = gen[0]
assert y.shape == (1, 3)
assert np.sum(y) == 3
def generator_flow(
self,
G,
link_ids,
link_targets,
sparse=False,
method="none",
k=1,
teleport_probability=0.1,
):
generator = FullBatchLinkGenerator(
G,
sparse=sparse,
method=method,
k=k,
teleport_probability=teleport_probability,
)
n_nodes = G.number_of_nodes()
gen = generator.flow(link_ids, link_targets)
if sparse:
[X, tind, A_ind, A_val], y = gen[0]
A_sparse = sps.coo_matrix(
(A_val[0], (A_ind[0, :, 0], A_ind[0, :, 1])),
shape=(n_nodes, n_nodes))
A_dense = A_sparse.toarray()
else:
[X, tind, A], y = gen[0]
A_dense = A[0]
assert np.allclose(X,
gen.features) # X should be equal to gen.features
assert isinstance(tind, np.ndarray)
assert tind.ndim == 3
assert tind.shape[1] == len(link_ids)
assert tind.shape[2] == 2
if link_targets is not None:
assert np.allclose(y, link_targets)
# Check that the diagonals are one
if method == "self_loops":
assert np.allclose(A_dense.diagonal(), 1)
return A_dense, tind, y
def test_APPNP_linkmodel_apply_dense():
G, features = create_graph_features()
adj = G.to_adjacency_matrix()
adj = np.array(adj.todense()[None, :, :])
generator = FullBatchLinkGenerator(G, sparse=False, method="none")
appnpnModel = APPNP([3], generator, activations=["relu"], dropout=0.5)
x_in, x_out = appnpnModel.in_out_tensors()
model = keras.Model(inputs=x_in, outputs=x_out)
# Check fit method
out_indices = np.array([[[0, 1], [1, 2]]], dtype="int32")
preds_1 = model.predict([features[None, :, :], out_indices, adj])
assert preds_1.shape == (1, 2, 2, 3)
# Check fit method
preds_2 = model.predict(generator.flow([("a", "b"), ("b", "c")]))
assert preds_2.shape == (1, 2, 2, 3)
assert preds_1 == pytest.approx(preds_2)
def test_APPNP_linkmodel_apply_sparse():
G, features = create_graph_features()
adj = nx.to_scipy_sparse_matrix(G)
features, adj = GCN_Aadj_feats_op(features, adj)
adj = adj.tocoo()
A_indices = np.expand_dims(np.hstack((adj.row[:, None], adj.col[:, None])),
0)
A_values = np.expand_dims(adj.data, 0)
nodes = G.nodes()
node_features = pd.DataFrame.from_dict(
{n: f
for n, f in zip(nodes, features)}, orient="index")
G = StellarGraph(G, node_features=node_features)
generator = FullBatchLinkGenerator(G, sparse=True, method="gcn")
appnpnModel = APPNP(layer_sizes=[3],
activations=["relu"],
generator=generator,
dropout=0.5)
x_in, x_out = appnpnModel.build()
model = keras.Model(inputs=x_in, outputs=x_out)
# Check fit method
out_indices = np.array([[[0, 1], [1, 2]]], dtype="int32")
preds_1 = model.predict(
[features[None, :, :], out_indices, A_indices, A_values])
assert preds_1.shape == (1, 2, 2, 3)
# Check fit_generator method
preds_2 = model.predict_generator(generator.flow([("a", "b"), ("b", "c")]))
assert preds_2.shape == (1, 2, 2, 3)
assert preds_1 == pytest.approx(preds_2)
# reduced graph G_test with the sampled links removed:
G_test, edge_ids_test, edge_labels_test = edge_splitter_test.train_test_split(
p=0.1, method="global", keep_connected=True)
# Define an edge splitter on the reduced graph G_test:
edge_splitter_train = EdgeSplitter(G_test)
# Randomly sample a fraction p=0.1 of all positive links, and same number of negative links, from G_test, and obtain the
# reduced graph G_train with the sampled links removed:
G_train, edge_ids_train, edge_labels_train = edge_splitter_train.train_test_split(
p=0.1, method="global", keep_connected=True)
epochs = 50
train_gen = FullBatchLinkGenerator(G_train, method="gcn")
train_flow = train_gen.flow(edge_ids_train, edge_labels_train)
test_gen = FullBatchLinkGenerator(G_test, method="gcn")
test_flow = train_gen.flow(edge_ids_test, edge_labels_test)
gcn = GCN(layer_sizes=[16, 16],
activations=["relu", "relu"],
generator=train_gen,
dropout=0.3)
x_inp, x_out = gcn.in_out_tensors()
prediction = LinkEmbedding(activation="relu", method="ip")(x_out)
prediction = keras.layers.Reshape((-1, ))(prediction)
def test_generator_flow_targets_not_iterator(self):
generator = FullBatchLinkGenerator(self.G)
link_ids = list(self.G.edges())[:3]
link_targets = 1
with pytest.raises(TypeError):
generator.flow(link_ids, link_targets)