[Misc] Always use subgraph instead of sub_graph. (#6213)

python/dgl/graphbolt/impl/neighbor_sampler.py

@@ -88,7 +88,7 @@ class NeighborSampler(SubgraphSampler):
         self.prob_name = prob_name
         self.graph = graph
 
-    def _sample_sub_graphs(self, seeds):
+    def _sample_subgraphs(self, seeds):
         subgraphs = []
         num_layers = len(self.fanouts)
         for hop in range(num_layers):

python/dgl/graphbolt/subgraph_sampler.py

@@ -40,9 +40,7 @@ class SubgraphSampler(Mapper):
             seeds = data.seed_node
         else:
             raise TypeError(f"Unsupported type of data {data}.")
-        data.input_nodes, data.sampled_subgraphs = self._sample_sub_graphs(
-            seeds
-        )
+        data.input_nodes, data.sampled_subgraphs = self._sample_subgraphs(seeds)
         return data
 
     def _link_prediction_preprocess(self, data):
@@ -105,5 +103,5 @@ class SubgraphSampler(Mapper):
             compacted_negative_tail if has_neg_dst else None,
         )
 
-    def _sample_sub_graphs(self, seeds):
+    def _sample_subgraphs(self, seeds):
         raise NotImplementedError