def test_walker_custom(line_graph):
walker = CustomWalker()
sampler = UnsupervisedSampler(line_graph, walker=walker)
batches = sampler.run(2)
assert len(batches) == line_graph.number_of_nodes()
# all positive examples should be self loops, since we defined our custom walker this way
for context_pairs, labels in batches:
for node, neighbour in context_pairs[labels == 1]:
assert node == neighbour
def test_run_batch_sizes(self, line_graph):
batch_size = 4
sampler = UnsupervisedSampler(G=line_graph, length=2, number_of_walks=2)
batches = sampler.run(batch_size)
# check batch sizes
assert len(batches) == np.ceil(len(line_graph.nodes()) * 4 / batch_size)
for ids, labels in batches[:-1]:
assert len(ids) == len(labels) == batch_size
# last batch can be smaller
ids, labels = batches[-1]
assert len(ids) == len(labels)
assert len(ids) <= batch_size
def test_ignored_param_warning(line_graph):
walker = UniformRandomWalk(line_graph, n=2, length=3)
with pytest.raises(ValueError,
match="cannot specify both 'walker' and 'length'"):
UnsupervisedSampler(line_graph, walker=walker, length=5)
with pytest.raises(
ValueError,
match="cannot specify both 'walker' and 'number_of_walks'"):
UnsupervisedSampler(line_graph, walker=walker, number_of_walks=5)
with pytest.raises(ValueError,
match="cannot specify both 'walker' and 'seed'"):
UnsupervisedSampler(line_graph, walker=walker, seed=1)
def test_walker_uniform_random(line_graph):
length = 3
number_of_walks = 2
batch_size = 4
walker = UniformRandomWalk(line_graph, n=number_of_walks, length=length)
sampler = UnsupervisedSampler(line_graph, walker=walker)
batches = sampler.run(batch_size)
# batches should match the parameters used to create the walker object, instead of the defaults
# for UnsupervisedSampler
expected_num_batches = np.ceil(line_graph.number_of_nodes() *
number_of_walks * (length - 1) * 2 /
batch_size)
assert len(batches) == expected_num_batches
def test_generator_multiple_batches(self):
n_feat = 4
batch_size = 4
number_of_batches = 3
G = example_Graph_2(n_feat)
sampler = UnsupervisedSampler(G=G)
sample_gen = sampler.generator(batch_size)
batches = []
for batch in range(number_of_batches):
batches.append(next(sample_gen))
assert len(batches) == number_of_batches
def test_UnsupervisedSampler_parameter(self):
g = create_test_graph()
# rw = UniformRandomWalk(StellarGraph(g))
# if no graph is provided
with pytest.raises(ValueError):
UnsupervisedSampler(G=None)
# graph has to be a Stellargraph object
with pytest.raises(ValueError):
UnsupervisedSampler(G=g)
g = StellarGraph(g)
"""
# only Uniform random walk is supported at the moment
with pytest.raises(TypeError):
UnsupervisedSampler(G=g, walker="any random walker")
# if no walker is provided, default to Uniform Random Walk
sampler = UnsupervisedSampler(G=g)
assert isinstance(sampler.walker, UniformRandomWalk)
"""
# walk must have length strictly greater than 1
with pytest.raises(ValueError):
UnsupervisedSampler(G=g, length=1)
# at least 1 walk from each root node
with pytest.raises(ValueError):
UnsupervisedSampler(G=g, number_of_walks=0)
# nodes nodes parameter should be an iterable of node IDs
with pytest.raises(ValueError):
UnsupervisedSampler(G=g, nodes=1)
# if no root nodes are provided for sampling defaulting to using all nodes as root nodes
sampler = UnsupervisedSampler(G=g, nodes=None)
assert sampler.nodes == list(g.nodes())
# if the seed value is provided check
# that the random choices is reproducable
sampler = UnsupervisedSampler(G=g, seed=1)
assert sampler.random.choices(range(100), k=10) == [
13,
84,
76,
25,
49,
44,
65,
78,
9,
2,
]
def test_run_context_pairs(self, line_graph):
batch_size = 4
sampler = UnsupervisedSampler(G=line_graph, length=2, number_of_walks=2)
batches = sampler.run(batch_size)
grouped_by_target = defaultdict(list)
for ids, labels in batches:
for (target, context), label in zip(ids, labels):
grouped_by_target[target].append((context, label))
assert len(grouped_by_target) == len(line_graph.nodes())
for target, sampled in grouped_by_target.items():
# exactly 2 positive and 2 negative context pairs for each target node
assert len(sampled) == 4
# since each walk has length = 2, there must be an edge between each positive context pair
for context, label in sampled:
if label == 1:
assert context in set(line_graph.neighbors(target))
def test_UnsupervisedSampler_parameter(self, line_graph):
# if no graph is provided
with pytest.raises(ValueError):
UnsupervisedSampler(G=None)
# walk must have length strictly greater than 1
with pytest.raises(ValueError):
UnsupervisedSampler(G=line_graph, length=1)
# at least 1 walk from each root node
with pytest.raises(ValueError):
UnsupervisedSampler(G=line_graph, number_of_walks=0)
# nodes nodes parameter should be an iterable of node IDs
with pytest.raises(ValueError):
UnsupervisedSampler(G=line_graph, nodes=1)
# if no root nodes are provided for sampling defaulting to using all nodes as root nodes
sampler = UnsupervisedSampler(G=line_graph, nodes=None)
assert sampler.nodes == list(line_graph.nodes())
def test_generator_samples(self):
n_feat = 4
batch_size = 4
G = example_Graph_2(n_feat)
sampler = UnsupervisedSampler(G=G)
sample_gen = sampler.generator(batch_size)
samples = next(sample_gen)
# return two lists: [(target,context)] pairs and [1/0] binary labels
assert len(samples) == 2
# each (target, context) pair has a matching label
assert len(samples[0]) == len(samples[1])
# batch-size number of samples are returned if batch_size is even
assert len(samples[0]) == batch_size
def unsup_gs(
g,
num_samples,
optimizer,
batch_size=4,
epochs=4,
bias=True,
dropout=0.0,
normalize="l2",
number_of_walks=1,
walk_length=5,
seed=0,
shuffle=True,
):
set_seed(seed)
tf.random.set_seed(seed)
if shuffle:
random.seed(seed)
nodes = list(g.nodes())
unsupervised_samples = UnsupervisedSampler(g,
nodes=nodes,
length=walk_length,
number_of_walks=number_of_walks)
generator = GraphSAGELinkGenerator(g, batch_size, num_samples)
train_gen = generator.flow(unsupervised_samples)
model = unsup_gs_model(num_samples, generator, optimizer, bias, dropout,
normalize)
model.fit(
train_gen,
epochs=epochs,
verbose=1,
use_multiprocessing=False,
workers=4,
shuffle=shuffle,
)
return model
def test_UnsupervisedSampler_parameter(self, line_graph):
# if no graph is provided
with pytest.raises(ValueError):
UnsupervisedSampler(G=None)
# walk must have length strictly greater than 1
with pytest.raises(ValueError):
UnsupervisedSampler(G=line_graph, length=1)
# at least 1 walk from each root node
with pytest.raises(ValueError):
UnsupervisedSampler(G=line_graph, number_of_walks=0)
# nodes nodes parameter should be an iterable of node IDs
with pytest.raises(ValueError):
UnsupervisedSampler(G=line_graph, nodes=1)
# if no root nodes are provided for sampling defaulting to using all nodes as root nodes
sampler = UnsupervisedSampler(G=line_graph, nodes=None)
assert sampler.nodes == list(line_graph.nodes())
# if the seed value is provided check
# that the random choices is reproducable
sampler = UnsupervisedSampler(G=line_graph, seed=1)
assert sampler.random.choices(range(100), k=10) == [
13,
84,
76,
25,
49,
44,
65,
78,
9,
2,
]
def test_generator_parameter(self):
g = create_test_graph()
g = StellarGraph(g)
# rw = UniformRandomWalk(g)
sampler = UnsupervisedSampler(G=g)
# generator should be provided with a valid batch size. i.e. an integer >=1
sample_gen = sampler.generator(batch_size=None)
with pytest.raises(ValueError):
next(sample_gen)
sample_gen = sampler.generator(batch_size="x")
with pytest.raises(TypeError):
next(sample_gen)
sample_gen = sampler.generator(batch_size=0)
with pytest.raises(ValueError):
next(sample_gen)
sample_gen = sampler.generator(batch_size=3)
with pytest.raises(ValueError):
next(sample_gen)
