Add Graph Profiler Update to Profile Builder

dataprofiler/profilers/graph_profiler.py

@@ -6,8 +6,7 @@
 import pandas as pd
 import scipy.stats as st
 
-from dataprofiler.data_readers.graph_data import GraphData
-
+from ..data_readers.graph_data import GraphData
 from . import BaseColumnProfiler
 
 
@@ -19,16 +18,15 @@ class GraphProfiler(object):
     Statistical properties of graph
     """
 
-    def __init__(self, name, options=None):
+    def __init__(self, data, options=None):
         """
         Initialize Graph Profiler.
 
-        :param name: Name of the data
+        :param data: data
         :type name: String
         :param options: Options for the Graph Profiler
         :type options: GraphOptions
         """
-        self.name = name
         self.sample_size = 0
         self.times = defaultdict(float)
 
@@ -153,7 +151,7 @@ def update(self, graph):
         if isinstance(graph, GraphData):
             graph = graph.data
         graph_size = graph.size()
-        if graph_size == 0:
+        if graph_size == 0 and graph.number_of_nodes() == 0:
             return self
         profile = dict(sample_size=graph_size)
 

dataprofiler/profilers/profile_builder.py

@@ -2609,7 +2609,9 @@ def __new__(
 
         # Construct based off of initial kwarg input or inference
         if profiler_type == "graph":
-            return GraphProfiler(data, options=options)
+            profile = GraphProfiler(data, options=options)
+            profile.update(data)
+            return profile
         elif profiler_type == "structured":
             return StructuredProfiler(
                 data, samples_per_update, min_true_samples, options

dataprofiler/tests/data_readers/test_csv_graph_data.py

@@ -184,7 +184,7 @@ def test_data_loader_nodes(self):
             if input_file["list_nodes"] is None:
                 continue
             data = GraphData(
-                input_file_path=input_file["path"], data=None, options=options
+                input_file_path=input_file["path"], options=options
             )
             self.assertEqual(input_file["list_nodes"], sorted(data.nodes))
 
@@ -201,7 +201,7 @@ def test_data_loader_edges(self):
             if input_file["list_edges"] is None:
                 continue
             data = GraphData(
-                input_file_path=input_file["path"], data=None, options=options
+                input_file_path=input_file["path"], options=options
             )
             data_edges = list(data.edges)
 

dataprofiler/tests/profilers/test_graph_profiler.py

@@ -111,6 +111,7 @@ def test_report(self):
     def test_graph_data_object(self):
         data = GraphData(input_file_path=None, data=self.graph)
         graph_profile = GraphProfiler("test_graph_data_object_update")
+
         with utils.mock_timeit():
             profile = graph_profile.update(data)
         scale = profile.profile["continuous_distribution"]["weight"].pop("scale")

dataprofiler/tests/profilers/test_profile_builder.py

@@ -3622,10 +3622,6 @@ def test_profiler_factory_class_bad_input(self):
         "dataprofiler.profilers.profile_builder.UnstructuredProfiler",
         spec=UnstructuredProfiler,
     )
-    @mock.patch(
-        "dataprofiler.profilers.graph_profiler.GraphProfiler",
-        spec=GraphProfiler,
-    )
     def test_profiler_factory_class_creates_correct_profiler(self, *mocks):
         """
         Ensure Profiler factory class either respects user input or makes
@@ -3670,6 +3666,12 @@ def test_profiler_factory_class_creates_correct_profiler(self, *mocks):
         data_str = "test"
         self.assertIsInstance(Profiler(data_str), UnstructuredProfiler)
 
+        data_graph.add_node(1)
+        data_graph = dp.Data(data=data_graph, data_type="graph")
+        graph_profile = Profiler(data_graph)
+        profile = graph_profile.profile
+        self.assertIsNotNone(profile.get("num_nodes"))
+
     def test_save_and_load_structured(self):
         datapth = "dataprofiler/tests/data/"
         test_files = ["csv/guns.csv", "csv/iris.csv"]