def jaccard_similarity_kg(graph: KatanaGraph,
compare_node: NodeID) -> NumpyVectorType:
jaccard_prop_name = "jaccard_prop_with_" + str(compare_node)
if not has_node_prop(graph.value, jaccard_prop_name):
jaccard(graph.value, compare_node, jaccard_prop_name)
jaccard_similarities = graph.value.get_node_property(
jaccard_prop_name).to_numpy()
return jaccard_similarities
def test_jaccard_sorted(graph: Graph):
sort_all_edges_by_dest(graph)
property_name = "NewProp"
compare_node = 0
jaccard(graph, compare_node, property_name, JaccardPlan.sorted())
jaccard_assert_valid(graph, compare_node, property_name)
similarities: np.ndarray = graph.get_node_property(
property_name).to_numpy()
assert similarities[compare_node] == 1
assert similarities[1917] == approx(0.0)
assert similarities[2812] == approx(0.0)
def test_jaccard(graph: Graph):
property_name = "NewProp"
compare_node = 0
jaccard(graph, compare_node, property_name)
node_schema: Schema = graph.loaded_node_schema()
num_node_properties = len(node_schema)
new_property_id = num_node_properties - 1
assert node_schema.names[new_property_id] == property_name
jaccard_assert_valid(graph, compare_node, property_name)
stats = JaccardStatistics(graph, compare_node, property_name)
assert stats.max_similarity == approx(1)
assert stats.min_similarity == approx(0)
assert stats.average_similarity == approx(0.000552534)
similarities: np.ndarray = graph.get_node_property(
property_name).to_numpy()
assert similarities[compare_node] == 1
assert similarities[1917] == approx(0.0)
assert similarities[2812] == approx(0.0)