Added utility operation for getting unannotated columns

dlhub_sdk/models/datasets.py

@@ -103,6 +103,11 @@ def annotate_column(self, column_name, description=None, data_type=None, units=N
             column['units'] = units
         return self
 
+    def get_unannotated_columns(self):
+        """Get the names of columns that have not been described"""
+
+        return [x["name"] for x in self["dataset"]["columns"] if "description" not in x]
+
     def _get_column(self, column_name):
         """Gets the metadata for a certain column
 

dlhub_sdk/models/tests/test_datasets.py

@@ -190,6 +190,16 @@ def test_serialize(self):
         # Make sure the fields are the same
         self.assertEqual(metadata._output, metadata_copy._output)
 
+    def test_unlabeled(self):
+        data_path = os.path.abspath(os.path.join(os.path.dirname(__file__), 'test.csv'))
+        m = TabularDataset.create_model(data_path)
+
+        # Check that all columns are unlabeled
+        self.assertEqual(["x", "y"], m.get_unannotated_columns())
+
+        # Label one, make sure it changes
+        m.annotate_column("x", "description")
+        self.assertEqual(["y"], m.get_unannotated_columns())
 
 if __name__ == "__main__":
     unittest.main()