def test_project_subgraph(arrow_modern_graph):
graph = arrow_modern_graph
sub_graph = graph.project(vertices={}, edges={})
assert sub_graph.schema.vertex_labels == []
assert sub_graph.schema.edge_labels == []
with pytest.raises(
RuntimeError,
match="Failed to project to simple graph as no vertex exists in this graph",
):
graphscope.wcc(sub_graph)
# project a sub_graph only contain person nodes
sub_graph = graph.project(vertices={"person": None}, edges={})
assert sub_graph.schema.vertex_labels == ["person"]
assert sub_graph.schema.edge_labels == []
with pytest.raises(
RuntimeError,
match="Failed to project to simple graph as no edge exists in this graph",
):
graphscope.wcc(sub_graph)
graph = graph.project(
vertices={"person": None, "software": ["name", "id"]},
edges={"created": ["eid", "weight"], "knows": None},
)
assert graph.schema.vertex_labels == ["person", "software"]
assert graph.schema.edge_labels == ["knows", "created"]
assert [p.id for p in graph.schema.get_vertex_properties("person")] == [0, 1, 2]
assert [p.id for p in graph.schema.get_vertex_properties("software")] == [0, 2]
assert [p.id for p in graph.schema.get_edge_properties("created")] == [0, 1]
assert [p.id for p in graph.schema.get_edge_properties("knows")] == [0, 1]
graph = graph.project(edges={"knows": ["eid"]}, vertices={"person": None})
assert graph.schema.vertex_labels == ["person"]
assert graph.schema.edge_labels == ["knows"]
assert [p.id for p in graph.schema.get_vertex_properties("person")] == [0, 1, 2]
assert [p.id for p in graph.schema.get_edge_properties("knows")] == [0]
with pytest.raises(ValueError, match="weight not exist in properties"):
graph = graph.project(edges={"knows": ["weight"]}, vertices={"person": []})
graph = graph.project(edges={"knows": []}, vertices={"person": []})
assert not graph.schema.get_vertex_properties("person")
assert not graph.schema.get_edge_properties("knows")
ret = graphscope.wcc(graph)
graph = graph.add_column(ret, {"cc": "r"})
assert len(graph.schema.get_vertex_properties("person")) == 1
assert graph.schema.get_vertex_properties("person")[0].name == "cc"
def test_add_column(ldbc_graph, arrow_modern_graph):
ldbc = ldbc_graph
modern = arrow_modern_graph
sub_graph_1 = ldbc.project(vertices={"person": []}, edges={"knows": ["eid"]})
sub_graph_2 = ldbc.project(
vertices={"person": None, "tag": None}, edges={"hasInterest": None}
)
sub_graph_3 = ldbc.project(
vertices={"tag": None, "tagclass": None}, edges={"hasType": None}
)
sub_graph_4 = modern.project(vertices={"person": []}, edges={"knows": ["eid"]})
ret = graphscope.wcc(sub_graph_1)
# the ret can add to the graph queried on
g1 = sub_graph_1.add_column(ret, selector={"cc": "r"})
assert g1.schema.get_vertex_properties("person")[0].id == 8
assert g1.schema.get_vertex_properties("person")[0].name == "cc"
# the ret can add to the origin graph
g2 = ldbc.add_column(ret, selector={"cc": "r"})
assert g2.schema.get_vertex_properties("person")[8].id == 8
assert g2.schema.get_vertex_properties("person")[8].name == "cc"
# the ret can add to the graph tha contain the same vertex label with sub_graph_1
g3 = sub_graph_2.add_column(ret, selector={"cc": "r"})
assert g3.schema.get_vertex_properties("person")[8].id == 8
assert g3.schema.get_vertex_properties("person")[8].name == "cc"
# the ret can not add to sub_graph_3
with pytest.raises(AnalyticalEngineInternalError):
g4 = sub_graph_3.add_column(ret, selector={"cc": "r"})
print(g4.schema)
# the ret can not add to sub_graph_4
with pytest.raises(AnalyticalEngineInternalError):
g5 = sub_graph_4.add_column(ret, selector={"cc": "r"})
print(g4.schema)
def test_app_on_undirected_graph(
p2p_project_undirected_graph,
sssp_result,
pagerank_result,
bfs_result,
wcc_result,
lpa_result,
triangles_result,
kshell_result,
):
# sssp
ctx1 = sssp(p2p_project_undirected_graph, src=6)
r1 = (ctx1.to_dataframe({
"node": "v.id",
"r": "r"
}).sort_values(by=["node"]).to_numpy(dtype=float))
r1[r1 == 1.7976931348623157e308] = float(
"inf") # replace limit<double>::max with inf
assert np.allclose(r1, sssp_result["undirected"])
assert np.allclose(
ctx1.to_dataframe({
"node": "v.id",
"r": "r"
},
vertex_range={
"begin": 1,
"end": 4
}).sort_values(by=["node"]).to_numpy(),
[[1.0, 31.0], [2.0, 39.0], [3.0, 78.0]],
)
assert np.allclose(
sorted(ctx1.to_numpy("r", vertex_range={
"begin": 1,
"end": 4
})),
[31.0, 39.0, 78.0],
)
# pagerank (only work on undirected graph)
ctx2 = pagerank(p2p_project_undirected_graph, delta=0.85, max_round=10)
r2 = (ctx2.to_dataframe({
"node": "v.id",
"r": "r"
}).sort_values(by=["node"]).to_numpy(dtype=float))
assert np.allclose(r2, pagerank_result["undirected"])
ctx3 = pagerank(p2p_project_undirected_graph, 0.85, 10)
r3 = (ctx3.to_dataframe({
"node": "v.id",
"r": "r"
}).sort_values(by=["node"]).to_numpy(dtype=float))
assert np.allclose(r3, pagerank_result["undirected"])
# r4 = pagerank(arrow_project_graph, 10, 0.85) # check max_round=10
# assert r4 is not None
ctx5 = pagerank(p2p_project_undirected_graph, "0.85", "10")
r5 = (ctx5.to_dataframe({
"node": "v.id",
"r": "r"
}).sort_values(by=["node"]).to_numpy(dtype=float))
assert np.allclose(r5, pagerank_result["undirected"])
ctx6 = pagerank(p2p_project_undirected_graph)
r6 = (ctx6.to_dataframe({
"node": "v.id",
"r": "r"
}).sort_values(by=["node"]).to_numpy(dtype=float))
assert np.allclose(r6, pagerank_result["undirected"])
assert np.allclose(
ctx6.to_dataframe({
"node": "v.id",
"r": "r"
},
vertex_range={
"begin": 1,
"end": 4
}).sort_values(by=["node"]).to_numpy(),
[
[1.0, 6.153724343761569e-05],
[2.0, 9.280361872165397e-05],
[3.0, 1.643246086005906e-05],
],
)
assert np.allclose(
sorted(ctx6.to_numpy("r", vertex_range={
"begin": 1,
"end": 4
})),
sorted([
6.153724343761569e-05, 9.280361872165397e-05, 1.643246086005906e-05
]),
)
# bfs
ctx7 = bfs(p2p_project_undirected_graph, src=6)
r7 = (ctx7.to_dataframe({
"node": "v.id",
"r": "r"
}).sort_values(by=["node"]).to_numpy(dtype=int))
assert np.all(r7 == bfs_result["undirected"])
assert np.all(
ctx7.to_dataframe(
{
"node": "v.id",
"r": "r"
}, vertex_range={
"begin": 1,
"end": 4
}).sort_values(by=["node"]).to_numpy() == [[1, 1], [2, 2], [3, 2]])
assert np.all(
sorted(ctx7.to_numpy("r", vertex_range={
"begin": 1,
"end": 4
})) == [1, 2, 2])
# wcc
ctx8 = wcc(p2p_project_undirected_graph)
r8 = (ctx8.to_dataframe({
"node": "v.id",
"r": "r"
}).sort_values(by=["node"]).to_numpy(dtype=int))
assert np.all(r8 == wcc_result)
assert np.all(
ctx8.to_dataframe(
{
"node": "v.id",
"r": "r"
}, vertex_range={
"begin": 1,
"end": 4
}).sort_values(by=["node"]).to_numpy() == [[1, 1], [2, 1], [3, 1]])
assert np.all(
ctx8.to_numpy("r", vertex_range={
"begin": 1,
"end": 4
}) == [1, 1, 1])
# lpa
ctx9 = lpa(p2p_project_undirected_graph, max_round=10)
r9 = (ctx9.to_dataframe({
"node": "v.id",
"r": "r"
}).sort_values(by=["node"]).to_numpy(dtype=int))
assert np.all(r9 == lpa_result)
assert np.all(
ctx9.to_dataframe(
{
"node": "v.id",
"r": "r"
}, vertex_range={
"begin": 1,
"end": 4
}).sort_values(by=["node"]).to_numpy() == [[1, 1], [2, 2], [3, 2]])
assert np.all(
sorted(ctx9.to_numpy("r", vertex_range={
"begin": 1,
"end": 4
})) == [1, 2, 2])
# kshell
ctx10 = k_shell(p2p_project_undirected_graph, k=3)
r10 = (ctx10.to_dataframe({
"node": "v.id",
"r": "r"
}).sort_values(by=["node"]).to_numpy(dtype=int))
assert np.all(r10 == kshell_result)
assert np.all(
ctx10.to_dataframe(
{
"node": "v.id",
"r": "r"
}, vertex_range={
"begin": 1,
"end": 4
}).sort_values(by=["node"]).to_numpy() == [[1, 0], [2, 0], [3, 0]])
assert np.all(
ctx10.to_numpy("r", vertex_range={
"begin": 1,
"end": 4
}) == [0, 0, 0])
# triangles
ctx_triangles = triangles(p2p_project_undirected_graph)
ret_triangles = (ctx_triangles.to_dataframe({
"node": "v.id",
"r": "r"
}).sort_values(by=["node"]).to_numpy(dtype=float))
assert np.allclose(ret_triangles, triangles_result["undirected"])
# louvain
ctx10 = louvain(p2p_project_undirected_graph,
min_progress=50,
progress_tries=2)
# simple_path
assert is_simple_path(p2p_project_undirected_graph, [1, 10])
def test_graph_lifecycle(graphscope_session):
graph = load_modern_graph(graphscope_session)
c = graphscope.wcc(graph)
del graph
assert c.to_numpy("v.id") is not None
del c # should delete c and graph in c++