def test_cluster_stats(self):
X = np.random.random((1000, 3))
ids = np.random.choice(20, 1000)
cluster_data = visuals._format_cluster_statistics(ids, X, ["a", "b", "c"])
assert isinstance(cluster_data, dict)
assert cluster_data["size"] == len(ids)
def test_cluster_stats_sparse_coo(self):
X = scipy.sparse.random(1000, 3, density=1.0, format="coo")
ids = np.random.choice(20, 1000)
with pytest.raises(
ValueError,
match=r".*sparse matrix format must be csr or csc.*"):
cluster_data = visuals._format_cluster_statistics(
ids, X, ["a", "b", "c"])
def test_cluster_stats_sparse_csc(self):
X = scipy.sparse.random(1000, 3, density=1.0, format="csc")
ids = np.random.choice(20, 1000)
cluster_data = visuals._format_cluster_statistics(
ids, X, ["a", "b", "c"])
assert isinstance(cluster_data, dict)
assert cluster_data["size"] == len(ids)
def test_cluster_stats_below(self):
X = np.ones((1000, 3))
ids = np.random.choice(20, 1000)
X[ids, 0] = 0
cluster_data = visuals._format_cluster_statistics(ids, X, ["a", "b", "c"])
assert len(cluster_data["below"]) >= 1
assert cluster_data["below"][0]["feature"] == "a"
assert cluster_data["below"][0]["mean"] == 0
def test_cluster_stats_below(self):
X = np.ones((1000, 3))
ids = np.random.choice(20, 1000)
X[ids, 0] = 0
cluster_data = visuals._format_cluster_statistics(
ids, X, ["a", "b", "c"])
assert len(cluster_data['below']) >= 1
assert cluster_data['below'][0]['feature'] == 'a'
assert cluster_data['below'][0]['mean'] == 0
def test_cluster_stats_with_no_names(self):
# This would be the default.
X = np.ones((1000, 3))
ids = np.random.choice(20, 1000)
X[ids, 0] = 0
cluster_data = visuals._format_cluster_statistics(ids, X, [])
assert len(cluster_data["below"]) >= 1
assert cluster_data["below"][0]["feature"] == "f_0"
assert cluster_data["below"][0]["mean"] == 0
