diff --git a/python/dgl/distributed/partition.py b/python/dgl/distributed/partition.py
index 0bf6deaaba49..48005ffb4d27 100644
--- a/python/dgl/distributed/partition.py
+++ b/python/dgl/distributed/partition.py
@@ -1268,9 +1268,7 @@ def get_homogeneous(g, balance_ntypes):
                         int(F.as_scalar(inner_nids[-1])) + 1,
                     ]
                 )
-            val = np.cumsum(
-                val
-            ).tolist()  # note computing the cumulative sum of array elements.
+            val = np.cumsum(val).tolist()
             assert val[-1] == g.num_nodes(ntype)
         for etype in g.canonical_etypes:
             etype_id = g.get_etype_id(etype)
diff --git a/tools/distpartitioning/dataset_utils.py b/tools/distpartitioning/dataset_utils.py
index f6b542f34b53..8567988f192f 100644
--- a/tools/distpartitioning/dataset_utils.py
+++ b/tools/distpartitioning/dataset_utils.py
@@ -547,6 +547,8 @@ def get_dataset(
                     autogenerate_column_names=True,
                 )
                 parse_options = pyarrow.csv.ParseOptions(delimiter=" ")
+                # if getsize() == 0, the file is empty, indicating that the partition doesn't have this attribute.
+                # The src_ids and dst_ids should remain empty.
                 if os.path.getsize(edge_file) != 0:
                     with pyarrow.csv.open_csv(
                         edge_file,