Enable sampling multiple nodes per "example"

graphlearn_torch/python/sampler/base.py

@@ -60,10 +60,10 @@ def __getitem__(self, index: Union[torch.Tensor, Any]) -> 'NodeSamplerInput':
     if not isinstance(index, torch.Tensor):
       index = torch.tensor(index, dtype=torch.long)
     index = index.to(self.node.device)
-    return NodeSamplerInput(self.node[index], self.input_type)
+    return NodeSamplerInput(self.node[index].view(-1), self.input_type)
 
   def __len__(self):
-    return self.node.numel()
+    return self.node.shape[0]
 
   def share_memory(self):
     self.node.share_memory_()

test/python/test_sampler_base.py

@@ -0,0 +1,59 @@
+import unittest
+
+import torch
+from torch.testing import assert_close
+
+from graphlearn_torch.sampler.base import NodeSamplerInput
+
+
+def _assert_tensor_equal(tensor1, tensor2):
+    assert_close(tensor1, tensor2, rtol=0, atol=0)
+
+
+class TestSamplerBase(unittest.TestCase):
+
+    def test_node_sampler_input_int_index(self):
+        input_data = NodeSamplerInput(node=torch.arange(10))
+        _assert_tensor_equal(input_data[0].node, torch.tensor([0]))
+
+    def test_node_sampler_input_tensor_index(self):
+        input_data = NodeSamplerInput(node=torch.arange(10))
+        with self.subTest("scalar tensor input"):
+            _assert_tensor_equal(input_data[torch.tensor(0)].node, torch.tensor([0]))
+
+        with self.subTest("slice tensor input"):
+            _assert_tensor_equal(input_data[torch.tensor([0, 1])].node, torch.tensor([0, 1]))
+
+    def test_node_sampler_input_multiple_examples(self):
+        input_data = NodeSamplerInput(node=torch.tensor([[0, 1, 2], [3, 4, 5]]))
+        self.assertEqual(len(input_data), 2)
+
+        with self.subTest("scalar tensor input"):
+            _assert_tensor_equal(input_data[torch.tensor(0)].node, torch.tensor([0, 1, 2]))
+
+        with self.subTest("slice tensor input"):
+            _assert_tensor_equal(
+                input_data[torch.tensor([0, 1])].node, torch.tensor([0, 1, 2, 3, 4, 5])
+            )
+
+        # Also test with dataloader - since that's how we actually use this
+        with self.subTest("dataloader"):
+            loader = torch.utils.data.DataLoader(torch.arange(2))
+            expected_data = torch.tensor([[0, 1, 2], [3, 4, 5]])
+
+            self.assertEqual(len(loader), len(expected_data))
+            for data, expected in zip(loader, expected_data):
+                _assert_tensor_equal(input_data[data].node, expected)
+
+        # Also test with dataloader - since that's how we actually use this
+        with self.subTest("dataloader - batched"):
+            loader = torch.utils.data.DataLoader(torch.arange(2), batch_size=2)
+            expected_data = torch.tensor([[0, 1, 2, 3, 4, 5]])
+
+            self.assertEqual(len(loader), len(expected_data))
+            for data, expected in zip(loader, expected_data):
+                _assert_tensor_equal(input_data[data].node, expected)
+
+
+if __name__ == "__main__":
+    unittest.main()