[GraphBolt] Modify `_init_all_nodes_set` to use `seeds`. (#7243)

Co-authored-by: Ubuntu <ubuntu@ip-172-31-0-133.us-west-2.compute.internal>

examples/sampling/graphbolt/link_prediction.py

@@ -101,7 +101,7 @@ class SAGE(nn.Module):
                 if not is_last_layer:
                     hidden_x = F.relu(hidden_x)
                 # By design, our seed nodes are contiguous.
-                y[data.seed_nodes[0] : data.seed_nodes[-1] + 1] = hidden_x.to(
+                y[data.seeds[0] : data.seeds[-1] + 1] = hidden_x.to(
                     buffer_device, non_blocking=True
                 )
             if not is_last_layer:

examples/sampling/graphbolt/node_classification.py

@@ -218,7 +218,7 @@ class SAGE(nn.Module):
                     hidden_x = F.relu(hidden_x)
                     hidden_x = self.dropout(hidden_x)
                 # By design, our output nodes are contiguous.
-                y[data.seed_nodes[0] : data.seed_nodes[-1] + 1] = hidden_x.to(
+                y[data.seeds[0] : data.seeds[-1] + 1] = hidden_x.to(
                     buffer_device
                 )
             if not is_last_layer:

examples/sampling/graphbolt/pyg/node_classification_advanced.py

@@ -138,7 +138,7 @@ class GraphSAGE(torch.nn.Module):
                 if not is_last_layer:
                     hidden_x = F.relu(hidden_x)
                 # By design, our output nodes are contiguous.
-                y[data.seed_nodes[0] : data.seed_nodes[-1] + 1] = hidden_x.to(
+                y[data.seeds[0] : data.seeds[-1] + 1] = hidden_x.to(
                     buffer_device
                 )
             if not is_last_layer:

python/dgl/graphbolt/impl/ondisk_dataset.py

@@ -900,13 +900,13 @@ class OnDiskDataset(Dataset):
         if isinstance(num_nodes, int):
             return ItemSet(
                 torch.tensor(num_nodes, dtype=dtype),
-                names="seed_nodes",
+                names="seeds",
             )
         else:
             data = {
                 node_type: ItemSet(
                     torch.tensor(num_node, dtype=dtype),
-                    names="seed_nodes",
+                    names="seeds",
                 )
                 for node_type, num_node in num_nodes.items()
             }

tests/python/pytorch/graphbolt/gb_test_utils.py

@@ -198,21 +198,21 @@ def random_homo_graphbolt_graph(
             train_set:
               - type: null
                 data:
-                  - name: node_pairs
+                  - name: seeds
                     format: numpy
                     in_memory: true
                     path: {train_path}
             validation_set:
               - type: null
                 data:
-                  - name: node_pairs
+                  - name: seeds
                     format: numpy
                     in_memory: true
                     path: {validation_path}
             test_set:
               - type: null
                 data:
-                  - name: node_pairs
+                  - name: seeds
                     format: numpy
                     in_memory: true
                     path: {test_path}
@@ -349,21 +349,21 @@ def generate_raw_data_for_hetero_dataset(
             train_set:
               - type: user
                 data:
-                  - name: seed_nodes
+                  - name: seeds
                     format: numpy
                     in_memory: true
                     path: {train_path}
             validation_set:
               - type: user
                 data:
-                  - name: seed_nodes
+                  - name: seeds
                     format: numpy
                     in_memory: true
                     path: {validation_path}
             test_set:
               - type: user
                 data:
-                  - name: seed_nodes
+                  - name: seeds
                     format: numpy
                     in_memory: true
                     path: {test_path}

tests/python/pytorch/graphbolt/impl/test_ondisk_dataset.py

@@ -2326,7 +2326,7 @@ def test_OnDiskDataset_all_nodes_set_homo():
         dataset = write_yaml_and_load_dataset(yaml_content, test_dir)
         all_nodes_set = dataset.all_nodes_set
         assert isinstance(all_nodes_set, gb.ItemSet)
-        assert all_nodes_set.names == ("seed_nodes",)
+        assert all_nodes_set.names == ("seeds",)
         for i, item in enumerate(all_nodes_set):
             assert i == item
 
@@ -2365,7 +2365,7 @@ def test_OnDiskDataset_all_nodes_set_hetero():
         dataset = write_yaml_and_load_dataset(yaml_content, test_dir)
         all_nodes_set = dataset.all_nodes_set
         assert isinstance(all_nodes_set, gb.ItemSetDict)
-        assert all_nodes_set.names == ("seed_nodes",)
+        assert all_nodes_set.names == ("seeds",)
         for i, item in enumerate(all_nodes_set):
             assert len(item) == 1
             assert isinstance(item, dict)