def test_pickle_bad_pickle():
with assert_warns_message(
UserWarning,
"Returning object that is not a graphtools.base.BaseGraph"):
with tempfile.TemporaryDirectory() as tempdir:
path = os.path.join(tempdir, "tmp.pkl")
with open(path, "wb") as f:
pickle.dump("hello world", f)
G = graphtools.read_pickle(path)
def test_sample_idx_unique():
with assert_raises_message(
ValueError, "sample_idx must contain more than one unique value"):
build_graph(data,
graph_class=graphtools.graphs.MNNGraph,
sample_idx=np.ones(len(data)))
with assert_warns_message(UserWarning,
"Only one unique sample. Not using MNNGraph"):
build_graph(data, sample_idx=np.ones(len(data)), graphtype="mnn")
def test_precomputed_with_pca():
with assert_warns_message(
RuntimeWarning,
"n_pca cannot be given on a precomputed graph. Setting n_pca=None",
):
build_graph(squareform(pdist(data)),
precomputed="distance",
n_pca=20,
decay=10)
def test_from_igraph_invalid_attribute():
with assert_warns_message(
UserWarning,
"Edge attribute invalid not found. Returning unweighted graph"):
n = 100
m = 500
K = np.zeros((n, n))
for _ in range(m):
e = np.random.choice(n, 2, replace=False)
K[e[0], e[1]] = K[e[1], e[0]] = 1
g = igraph.Graph.Adjacency(K.tolist())
G = graphtools.from_igraph(g, attribute="invalid")
def test_knn_graph_sparse_no_pca():
with assert_warns_message(
UserWarning,
"cannot use tree with sparse input: using brute force"):
build_graph(
sp.coo_matrix(data),
n_pca=None, # n_pca,
decay=10,
knn=3,
thresh=1e-4,
random_state=42,
use_pygsp=True,
)
def test_from_igraph_invalid_precomputed():
with assert_warns_message(
UserWarning,
"Cannot build graph from igraph with precomputed=affinity. Use 'adjacency' instead.",
):
n = 100
m = 500
K = np.zeros((n, n))
for _ in range(m):
e = np.random.choice(n, 2, replace=False)
K[e[0], e[1]] = K[e[1], e[0]] = 1
g = igraph.Graph.Adjacency(K.tolist())
G = graphtools.from_igraph(g, attribute=None, precomputed="affinity")
def test_mnn_with_kernel_symmm_theta_and_no_theta():
with assert_warns_message(
UserWarning,
"kernel_symm='mnn' but theta not given. Defaulting to theta=1."):
build_graph(
data,
thresh=0,
n_pca=20,
decay=10,
knn=5,
random_state=42,
sample_idx=digits["target"],
kernel_symm="mnn",
)
def test_mnn_with_gamma():
with assert_warns_message(FutureWarning,
"gamma is deprecated. Setting theta=0.9"):
build_graph(
data,
thresh=0,
n_pca=20,
decay=10,
knn=5,
random_state=42,
sample_idx=digits["target"],
kernel_symm="mnn",
gamma=0.9,
)
def test_mnn_with_kernel_symm_theta():
with assert_warns_message(
FutureWarning,
"kernel_symm='theta' is deprecated. Setting kernel_symm='mnn'"):
build_graph(
data,
thresh=0,
n_pca=20,
decay=10,
knn=5,
random_state=42,
sample_idx=digits["target"],
kernel_symm="theta",
theta=0.9,
)
def test_mnn_with_theta_and_kernel_symm_not_theta():
with assert_warns_message(
UserWarning,
"kernel_symm='+' but theta is not None. Setting kernel_symm='mnn'."
):
build_graph(
data,
thresh=0,
n_pca=20,
decay=10,
knn=5,
random_state=42,
sample_idx=digits["target"],
kernel_symm="+",
theta=0.9,
)
def test_mnn_adaptive_k():
with assert_warns_message(
DeprecationWarning,
"`adaptive_k` has been deprecated. Using fixed knn."):
build_graph(
data,
thresh=0,
n_pca=20,
decay=10,
knn=5,
random_state=42,
sample_idx=digits["target"],
kernel_symm="mnn",
theta=0.9,
adaptive_k="sqrt",
)
def test_knn_graph():
k = 3
n_pca = 20
pca = PCA(n_pca, svd_solver="randomized", random_state=42).fit(data)
data_nu = pca.transform(data)
pdx = squareform(pdist(data_nu, metric="euclidean"))
knn_dist = np.partition(pdx, k, axis=1)[:, :k]
epsilon = np.max(knn_dist, axis=1)
K = np.empty_like(pdx)
for i in range(len(pdx)):
K[i, pdx[i, :] <= epsilon[i]] = 1
K[i, pdx[i, :] > epsilon[i]] = 0
K = K + K.T
W = np.divide(K, 2)
np.fill_diagonal(W, 0)
G = pygsp.graphs.Graph(W)
G2 = build_graph(data,
n_pca=n_pca,
decay=None,
knn=k - 1,
random_state=42,
use_pygsp=True)
assert G.N == G2.N
np.testing.assert_equal(G.dw, G2.dw)
assert (G.W - G2.W).nnz == 0
assert (G2.W - G.W).sum() == 0
assert isinstance(G2, graphtools.graphs.kNNGraph)
K2 = G2.build_kernel_to_data(G2.data_nu, knn=k)
K2 = (K2 + K2.T) / 2
assert (G2.K - K2).nnz == 0
assert (G2.build_kernel_to_data(
G2.data_nu, knn=data.shape[0]).nnz == data.shape[0] * data.shape[0])
with assert_warns_message(
UserWarning,
"Cannot set knn ({}) to be greater than "
"n_samples ({}). Setting knn={}".format(data.shape[0] + 1,
data.shape[0],
data.shape[0]),
):
G2.build_kernel_to_data(
Y=G2.data_nu,
knn=data.shape[0] + 1,
)
def test_if_sparse_deprecated():
with assert_warns_message(
DeprecationWarning,
"Call to deprecated function (or staticmethod) if_sparse. (Use graphtools.matrix.if_sparse instead) -- Deprecated since version 1.5.0.",
):
graphtools.utils.if_sparse(lambda x: x, lambda x: x, np.zeros((4, 4)))
def test_build_landmark_with_too_few_points():
with assert_warns_message(
RuntimeWarning,
"n_svd (100) >= n_samples (50) Consider using kNNGraph or lower n_svd",
):
build_graph(data[:50], n_landmark=25, n_svd=100)
def test_sparse_set_diagonal_deprecated():
with assert_warns_message(
DeprecationWarning,
"Call to deprecated function (or staticmethod) sparse_set_diagonal. (Use graphtools.matrix.sparse_set_diagonal instead) -- Deprecated since version 1.5.0.",
):
graphtools.utils.sparse_set_diagonal(sparse.csr_matrix((4, 4)), 1)
def test_single_sample_idx_warning():
with assert_warns_message(UserWarning,
"Only one unique sample. Not using MNNGraph"):
build_graph(data, sample_idx=np.repeat(1, len(data)))
def test_to_pygsp_invalid_use_pygsp():
with assert_warns_message(
UserWarning,
"Cannot build PyGSPGraph with use_pygsp=False. Use True instead."):
G = build_graph(data)
G2 = G.to_pygsp(use_pygsp=False)
def test_threshold_ignored():
with assert_warns_message(
RuntimeWarning,
"n_pca = 10, therefore rank_threshold of -1 will not be used. To use rank thresholding, set n_pca = True",
):
assert build_graph(data, n_pca=10, rank_threshold=-1).n_pca == 10
def test_fractional_n_pca():
with assert_warns_message(
RuntimeWarning, "Cannot perform PCA to fractional 1.5 dimensions. Rounding to 2"
):
build_graph(data, n_pca=1.5)
def test_knnmax_too_large():
with assert_warns_message(
UserWarning,
"Cannot set knn_max (9) to be less than knn (10). Setting knn_max=10",
):
build_graph(data, n_pca=20, decay=10, knn=10, knn_max=9, thresh=1e-4)
def test_to_array_deprecated():
with assert_warns_message(
DeprecationWarning,
"Call to deprecated function (or staticmethod) to_array. (Use graphtools.matrix.to_array instead) -- Deprecated since version 1.5.0.",
):
graphtools.utils.to_array([1])
def test_balltree_cosine():
with assert_warns_message(
UserWarning,
"Metric cosine not valid for `sklearn.neighbors.BallTree`. Graph instantiation may be slower than normal.",
):
build_graph(data, n_pca=20, decay=10, distance="cosine", thresh=1e-4)
def test_precomputed_nonzero_diagonal():
with assert_warns_message(RuntimeWarning,
"K should have a non-zero diagonal"):
build_graph(np.zeros((10, 10)), precomputed="affinity", n_pca=None)
def test_bandwidth_no_decay():
with assert_warns_message(UserWarning,
"`bandwidth` is not used when `decay=None`."):
build_graph(data, n_pca=20, decay=None, bandwidth=3, thresh=1e-4)
def test_nonzero_discrete_deprecated():
with assert_warns_message(
DeprecationWarning,
"Call to deprecated function (or staticmethod) nonzero_discrete. (Use graphtools.matrix.nonzero_discrete instead) -- Deprecated since version 1.5.0.",
):
graphtools.utils.nonzero_discrete(np.zeros((4, 4)), [1])
