Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Refine project_to_simple method. #186

Merged
merged 3 commits into from
Mar 15, 2021
+79 −84
Merged

Refine project_to_simple method. #186

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
1103881
Add checks for relationship when project
siyuan0322 Mar 12, 2021
c099e91
update
siyuan0322 Mar 12, 2021
b477419
pass
siyuan0322 Mar 12, 2021
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
14 changes: 8 additions & 6 deletions .github/workflows/ci.yml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -411,8 +411,8 @@ jobs:
- name: Clean
if: always()
run: |
sudo docker rmi registry.cn-hongkong.aliyuncs.com/graphscope/graphscope:${{ github.sha }} \
registry.cn-hongkong.aliyuncs.com/graphscope/maxgraph_standalone_manager:${{ github.sha }} || true
sudo docker rmi -f registry.cn-hongkong.aliyuncs.com/graphscope/graphscope:${{ github.sha }} \
registry.cn-hongkong.aliyuncs.com/graphscope/maxgraph_standalone_manager:${{ github.sha }} || true

gie-test:
runs-on: self-hosted
Expand Down Expand Up @@ -482,8 +482,8 @@ jobs:
- name: Clean
if: always()
run: |
sudo docker rmi registry.cn-hongkong.aliyuncs.com/graphscope/graphscope:${{ github.sha }} \
registry.cn-hongkong.aliyuncs.com/graphscope/maxgraph_standalone_manager:${{ github.sha }} || true
sudo docker rmi -f registry.cn-hongkong.aliyuncs.com/graphscope/graphscope:${{ github.sha }} \
registry.cn-hongkong.aliyuncs.com/graphscope/maxgraph_standalone_manager:${{ github.sha }} || true

release-image:
runs-on: ubuntu-20.04
Expand Down Expand Up @@ -526,5 +526,7 @@ jobs:
- name: Clean
if: always()
run: |
sudo docker rmi registry.cn-hongkong.aliyuncs.com/graphscope/graphscope:${{ github.sha }} \
registry.cn-hongkong.aliyuncs.com/graphscope/maxgraph_standalone_manager:${{ github.sha }} || true
sudo docker rmi -f registry.cn-hongkong.aliyuncs.com/graphscope/graphscope:latest \
registry.cn-hongkong.aliyuncs.com/graphscope/maxgraph_standalone_manager:latest \
registry.cn-hongkong.aliyuncs.com/graphscope/graphscope:${{ github.sha }} \
registry.cn-hongkong.aliyuncs.com/graphscope/maxgraph_standalone_manager:${{ github.sha }} || true
115 changes: 48 additions & 67 deletions python/graphscope/framework/graph.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -317,85 +317,64 @@ def project_to_simple(self, v_label="_", e_label="_", v_prop=None, e_prop=None):
raise RuntimeError(
"The graph is not registered in remote, and can't project to simple"
)
self.check_unmodified()
check_argument(self.graph_type == types_pb2.ARROW_PROPERTY)
check_argument(isinstance(v_label, (int, str)))
check_argument(isinstance(e_label, (int, str)))
self.check_unmodified()

def check_out_of_range(id, length):
if id < length and id > -1:
return id
else:
raise KeyError("id {} is out of range.".format(id))
if id >= length or id < 0:
raise IndexError("id {} is out of range.".format(id))

try:
v_label_id = (
check_out_of_range(v_label, self._schema.vertex_label_num)
if isinstance(v_label, int)
else self._schema.vertex_label_index(v_label)
)
if isinstance(v_label, str):
v_label_id = self._schema.vertex_label_index(v_label)
else:
v_label_id = v_label
check_out_of_range(v_label_id, self._schema.vertex_label_num)
v_label = self._schema.vertex_labels[v_label_id]
if isinstance(e_label, str):
e_label_id = self._schema.edge_label_index(e_label)
else:
e_label_id = e_label
check_out_of_range(e_label_id, self._schema.edge_label_num)
e_label = self._schema.edge_labels[e_label]
except ValueError as e:
raise ValueError("Label does not exists.") from e

# Check relation v_label -> e_label <- v_label exists.
relation = (v_label, v_label)
if relation not in self._schema.edge_relationships[e_label_id]:
raise ValueError(
"graph not contains the vertex label {}.".format(v_label)
) from e
f"Graph doesn't contain such relationship: {v_label} -> {e_label} <- {v_label}."
)

try:
e_label_id = (
check_out_of_range(e_label, self._schema.edge_label_num)
if isinstance(e_label, int)
else self._schema.edge_label_index(e_label)
)
if v_prop is None:
v_prop_id = -1
vdata_type = None
else:
if isinstance(v_prop, str):
v_prop_id = self._schema.vertex_property_index(v_label_id, v_prop)
else:
v_prop_id = v_prop
properties = self._schema.vertex_properties[v_label_id]
check_out_of_range(v_prop_id, len(properties))
vdata_type = list(properties.values())[v_prop_id]
if e_prop is None:
e_prop_id = -1
edata_type = None
else:
if isinstance(e_prop, str):
e_prop_id = self._schema.edge_property_index(e_label_id, e_prop)
else:
e_prop_id = e_prop
properties = self._schema.edge_properties[e_label_id]
check_out_of_range(e_prop_id, len(properties))
edata_type = list(properties.values())[e_prop_id]
except ValueError as e:
raise InvalidArgumentError(
"graph not contains the edge label {}.".format(e_label)
) from e

if v_prop is None:
# NB: -1 means vertex property is None
v_prop_id = -1
v_properties = None
else:
check_argument(isinstance(v_prop, (int, str)))
v_properties = self._schema.vertex_properties[v_label_id]
try:
v_prop_id = (
check_out_of_range(v_prop, len(v_properties))
if isinstance(v_prop, int)
else self._schema.vertex_property_index(v_label_id, v_prop)
)
except ValueError as e:
raise ValueError(
"vertex label {} not contains the property {}".format(
v_label, v_prop
)
) from e

if e_prop is None:
# NB: -1 means edge property is None
e_prop_id = -1
e_properties = None
else:
check_argument(isinstance(e_prop, (int, str)))
e_properties = self._schema.edge_properties[e_label_id]
try:
e_prop_id = (
check_out_of_range(e_prop, len(e_properties))
if isinstance(e_prop, int)
else self._schema.edge_property_index(e_label_id, e_prop)
)
except ValueError as e:
raise ValueError(
"edge label {} not contains the property {}".format(e_label, e_prop)
) from e
raise ValueError("Property does not exists.") from e

oid_type = self._schema.oid_type
vid_type = self._schema.vid_type
vdata_type = None
if v_properties:
vdata_type = list(v_properties.values())[v_prop_id]
edata_type = None
if e_properties:
edata_type = list(e_properties.values())[e_prop_id]

op = dag_utils.project_arrow_property_graph(
self,
Expand Down Expand Up @@ -424,8 +403,8 @@ def add_column(self, results, selector):
check_argument(
isinstance(selector, Mapping), "selector of add column must be a dict"
)
self.check_unmodified()
check_argument(self.graph_type == types_pb2.ARROW_PROPERTY)
self.check_unmodified()
selector = {
key: results._transform_selector(value) for key, value in selector.items()
}
Expand All @@ -443,6 +422,7 @@ def to_numpy(self, selector, vertex_range=None):
Returns:
`numpy.ndarray`
"""
check_argument(self.graph_type == types_pb2.ARROW_PROPERTY)
self.check_unmodified()
selector = utils.transform_labeled_vertex_property_data_selector(self, selector)
vertex_range = utils.transform_vertex_range(vertex_range)
Expand All @@ -460,6 +440,7 @@ def to_dataframe(self, selector, vertex_range=None):
Returns:
`pandas.DataFrame`
"""
check_argument(self.graph_type == types_pb2.ARROW_PROPERTY)
self.check_unmodified()
check_argument(
isinstance(selector, Mapping),
Expand Down
34 changes: 23 additions & 11 deletions python/tests/test_graph.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -84,41 +84,53 @@ def test_project_to_simple_with_id(p2p_property_graph, sssp_result):
def test_error_on_project_to_simple_id_out_of_range(arrow_property_graph):
g = arrow_property_graph
# g has 4 vertex labels and 2 edge labels, each label has 1 property
with pytest.raises(KeyError, match="id 5 is out of range"):
with pytest.raises(IndexError, match="id 5 is out of range"):
g.project_to_simple(v_label=5, e_label=0)

with pytest.raises(KeyError, match="id 3 is out of range"):
with pytest.raises(IndexError, match="id 3 is out of range"):
g.project_to_simple(v_label=0, e_label=3)

with pytest.raises(KeyError, match="id -1 is out of range"):
with pytest.raises(IndexError, match="id -1 is out of range"):
g.project_to_simple(v_label=-1, e_label=3)

with pytest.raises(KeyError, match="id 2 is out of range"):
with pytest.raises(IndexError, match="id 2 is out of range"):
g.project_to_simple(v_label=0, e_label=0, v_prop=2, e_prop=0)

with pytest.raises(KeyError, match="id 1 is out of range"):
with pytest.raises(IndexError, match="id 1 is out of range"):
g.project_to_simple(v_label=0, e_label=0, v_prop=0, e_prop=1)


def test_error_label_on_project_to_simple(arrow_property_graph):
g = arrow_property_graph
# g has vertex labels: v0, v1, v2, v3, each label has a property: weight
# g has edge label: e0, e1, each label has a property: dist
with pytest.raises(ValueError, match="graph not contains the vertex label v4"):
with pytest.raises(ValueError, match="Label does not exists"):
g.project_to_simple(v_label="v4", e_label="e0")

with pytest.raises(ValueError, match="graph not contains the edge label e2"):
with pytest.raises(ValueError, match="Label does not exists"):
g.project_to_simple(v_label="v0", e_label="e2")

with pytest.raises(
ValueError, match="vertex label v0 not contains the property foo"
):
with pytest.raises(ValueError, match="Property does not exists."):
g.project_to_simple(v_label="v0", e_label="e0", v_prop="foo")

with pytest.raises(ValueError, match="edge label e0 not contains the property foo"):
with pytest.raises(ValueError, match="Property does not exists."):
g.project_to_simple(v_label="v0", e_label="e0", e_prop="foo")


def test_error_relationship_on_project_to_simple(arrow_modern_graph):
g = arrow_modern_graph
with pytest.raises(
ValueError,
match="Graph doesn't contain such relationship: person -> created <- person",
):
g.project_to_simple(v_label="person", e_label="created")
with pytest.raises(
ValueError,
match="Graph doesn't contain such relationship: software -> knows <- software",
):
g.project_to_simple(v_label="software", e_label="knows")


def test_unload(graphscope_session):
prefix = os.path.expandvars("${GS_TEST_DIR}/property")
g = graphscope_session.load_from(
Expand Down