[DistGB] revert toindex() but refine tests (#7197)

python/dgl/distributed/dist_tensor.py

@@ -2,13 +2,12 @@
 
 import os
 
-from .. import backend as F
+from .. import backend as F, utils
 
 from .dist_context import is_initialized
 from .kvstore import get_kvstore
 from .role import get_role
 from .rpc import get_group_id
-from .utils import totensor
 
 
 def _default_init_data(shape, dtype):
@@ -201,11 +200,13 @@ class DistTensor:
             self.kvstore.delete_data(self._name)
 
     def __getitem__(self, idx):
-        idx = totensor(idx)
+        idx = utils.toindex(idx)
+        idx = idx.tousertensor()
         return self.kvstore.pull(name=self._name, id_tensor=idx)
 
     def __setitem__(self, idx, val):
-        idx = totensor(idx)
+        idx = utils.toindex(idx)
+        idx = idx.tousertensor()
         # TODO(zhengda) how do we want to support broadcast (e.g., G.ndata['h'][idx] = 1).
         self.kvstore.push(name=self._name, id_tensor=idx, data_tensor=val)
 

python/dgl/distributed/graph_partition_book.py

@@ -4,9 +4,8 @@ import pickle
 from abc import ABC
 
 import numpy as np
-import torch
 
-from .. import backend as F
+from .. import backend as F, utils
 from .._ffi.ndarray import empty_shared_mem
 from ..base import DGLError
 from ..ndarray import exist_shared_mem_array
@@ -762,6 +761,7 @@ class RangePartitionBook(GraphPartitionBook):
 
     def map_to_homo_nid(self, ids, ntype):
         """Map per-node-type IDs to global node IDs in the homogeneous format."""
+        ids = utils.toindex(ids).tousertensor()
         partids = self.nid2partid(ids, ntype)
         typed_max_nids = F.zerocopy_from_numpy(self._typed_max_node_ids[ntype])
         end_diff = F.gather_row(typed_max_nids, partids) - ids
@@ -772,6 +772,7 @@ class RangePartitionBook(GraphPartitionBook):
 
     def map_to_homo_eid(self, ids, etype):
         """Map per-edge-type IDs to global edge IDs in the homoenegeous format."""
+        ids = utils.toindex(ids).tousertensor()
         c_etype = self.to_canonical_etype(etype)
         partids = self.eid2partid(ids, c_etype)
         typed_max_eids = F.zerocopy_from_numpy(
@@ -785,28 +786,32 @@ class RangePartitionBook(GraphPartitionBook):
 
     def nid2partid(self, nids, ntype=DEFAULT_NTYPE):
         """From global node IDs to partition IDs"""
-        # [TODO][Rui] replace numpy with torch.
-        nids = nids.numpy()
+        nids = utils.toindex(nids)
         if ntype == DEFAULT_NTYPE:
-            ret = np.searchsorted(self._max_node_ids, nids, side="right")
+            ret = np.searchsorted(
+                self._max_node_ids, nids.tonumpy(), side="right"
+            )
         else:
             ret = np.searchsorted(
-                self._typed_max_node_ids[ntype], nids, side="right"
+                self._typed_max_node_ids[ntype], nids.tonumpy(), side="right"
             )
-        return torch.from_numpy(ret)
+        ret = utils.toindex(ret)
+        return ret.tousertensor()
 
     def eid2partid(self, eids, etype=DEFAULT_ETYPE):
         """From global edge IDs to partition IDs"""
-        # [TODO][Rui] replace numpy with torch.
-        eids = eids.numpy()
+        eids = utils.toindex(eids)
         if etype in (DEFAULT_ETYPE, DEFAULT_ETYPE[1]):
-            ret = np.searchsorted(self._max_edge_ids, eids, side="right")
+            ret = np.searchsorted(
+                self._max_edge_ids, eids.tonumpy(), side="right"
+            )
         else:
             c_etype = self.to_canonical_etype(etype)
             ret = np.searchsorted(
-                self._typed_max_edge_ids[c_etype], eids, side="right"
+                self._typed_max_edge_ids[c_etype], eids.tonumpy(), side="right"
             )
-        return torch.from_numpy(ret)
+        ret = utils.toindex(ret)
+        return ret.tousertensor()
 
     def partid2nids(self, partid, ntype=DEFAULT_NTYPE):
         """From partition ID to global node IDs"""
@@ -847,6 +852,8 @@ class RangePartitionBook(GraphPartitionBook):
                 getting remote tensor of nid2localnid."
             )
 
+        nids = utils.toindex(nids)
+        nids = nids.tousertensor()
         if ntype == DEFAULT_NTYPE:
             start = self._max_node_ids[partid - 1] if partid > 0 else 0
         else:
@@ -863,6 +870,8 @@ class RangePartitionBook(GraphPartitionBook):
                 getting remote tensor of eid2localeid."
             )
 
+        eids = utils.toindex(eids)
+        eids = eids.tousertensor()
         if etype in (DEFAULT_ETYPE, DEFAULT_ETYPE[1]):
             start = self._max_edge_ids[partid - 1] if partid > 0 else 0
         else:

python/dgl/distributed/graph_services.py

@@ -14,6 +14,7 @@ from ..sampling import (
     sample_neighbors as local_sample_neighbors,
 )
 from ..subgraph import in_subgraph as local_in_subgraph
+from ..utils import toindex
 from .rpc import (
     recv_responses,
     register_service,
@@ -707,6 +708,8 @@ def _distributed_access(g, nodes, issue_remote_req, local_access):
     """
     req_list = []
     partition_book = g.get_partition_book()
+    if not isinstance(nodes, torch.Tensor):
+        nodes = toindex(nodes).tousertensor()
     partition_id = partition_book.nid2partid(nodes)
     local_nids = None
     for pid in range(partition_book.num_partitions()):
@@ -900,7 +903,11 @@ def sample_etype_neighbors(
             ), "The sampled node type {} does not exist in the input graph".format(
                 ntype
             )
-            homo_nids.append(gpb.map_to_homo_nid(nodes[ntype], ntype))
+            if F.is_tensor(nodes[ntype]):
+                typed_nodes = nodes[ntype]
+            else:
+                typed_nodes = toindex(nodes[ntype]).tousertensor()
+            homo_nids.append(gpb.map_to_homo_nid(typed_nodes, ntype))
         nodes = F.cat(homo_nids, 0)
 
     def issue_remote_req(node_ids):
@@ -1025,7 +1032,11 @@ def sample_neighbors(
             assert (
                 ntype in g.ntypes
             ), "The sampled node type does not exist in the input graph"
-            homo_nids.append(gpb.map_to_homo_nid(nodes[ntype], ntype))
+            if F.is_tensor(nodes[ntype]):
+                typed_nodes = nodes[ntype]
+            else:
+                typed_nodes = toindex(nodes[ntype]).tousertensor()
+            homo_nids.append(gpb.map_to_homo_nid(typed_nodes, ntype))
         nodes = F.cat(homo_nids, 0)
     elif isinstance(nodes, dict):
         assert len(nodes) == 1
@@ -1095,6 +1106,7 @@ def _distributed_edge_access(g, edges, issue_remote_req, local_access):
     """
     req_list = []
     partition_book = g.get_partition_book()
+    edges = toindex(edges).tousertensor()
     partition_id = partition_book.eid2partid(edges)
     local_eids = None
     reorder_idx = []
@@ -1212,6 +1224,7 @@ def in_subgraph(g, nodes):
 def _distributed_get_node_property(g, n, issue_remote_req, local_access):
     req_list = []
     partition_book = g.get_partition_book()
+    n = toindex(n).tousertensor()
     partition_id = partition_book.nid2partid(n)
     local_nids = None
     reorder_idx = []
@@ -1256,21 +1269,7 @@ def _distributed_get_node_property(g, n, issue_remote_req, local_access):
 
 
 def in_degrees(g, v):
-    """Get in-degrees
-
-    Parameters
-    ----------
-
-    g : DistGraph
-        The distributed graph.
-    v : tensor
-        The node ID array.
-
-    Returns
-    -------
-    tensor
-        The in-degree array.
-    """
+    """Get in-degrees"""
 
     def issue_remote_req(v, order_id):
         return InDegreeRequest(v, order_id)
@@ -1282,21 +1281,7 @@ def in_degrees(g, v):
 
 
 def out_degrees(g, u):
-    """Get out-degrees
-
-    Parameters
-    ----------
-
-    g : DistGraph
-        The distributed graph.
-    u : tensor
-        The node ID array.
-
-    Returns
-    -------
-    tensor
-        The out-degree array.
-    """
+    """Get out-degrees"""
 
     def issue_remote_req(u, order_id):
         return OutDegreeRequest(u, order_id)

python/dgl/distributed/kvstore.py

@@ -4,7 +4,7 @@ import os
 
 import numpy as np
 
-from .. import backend as F
+from .. import backend as F, utils
 from .._ffi.ndarray import empty_shared_mem
 
 from . import rpc
@@ -1376,6 +1376,8 @@ class KVClient(object):
             a vector storing the global data ID
         """
         assert len(name) > 0, "name cannot be empty."
+        id_tensor = utils.toindex(id_tensor)
+        id_tensor = id_tensor.tousertensor()
         assert F.ndim(id_tensor) == 1, "ID must be a vector."
         # partition data
         machine_id = self._part_policy[name].to_partid(id_tensor)
@@ -1397,6 +1399,8 @@ class KVClient(object):
             a tensor with the same row size of data ID
         """
         assert len(name) > 0, "name cannot be empty."
+        id_tensor = utils.toindex(id_tensor)
+        id_tensor = id_tensor.tousertensor()
         assert F.ndim(id_tensor) == 1, "ID must be a vector."
         assert (
             F.shape(id_tensor)[0] == F.shape(data_tensor)[0]
@@ -1448,6 +1452,8 @@ class KVClient(object):
             a data tensor with the same row size of id_tensor.
         """
         assert len(name) > 0, "name cannot be empty."
+        id_tensor = utils.toindex(id_tensor)
+        id_tensor = id_tensor.tousertensor()
         assert F.ndim(id_tensor) == 1, "ID must be a vector."
         if self._pull_handlers[name] is default_pull_handler:  # Use fast-pull
             part_id = self._part_policy[name].to_partid(id_tensor)

python/dgl/distributed/nn/pytorch/sparse_emb.py

@@ -2,7 +2,7 @@
 
 import torch as th
 
-from .... import backend as F
+from .... import backend as F, utils
 from ...dist_tensor import DistTensor
 
 
@@ -99,7 +99,7 @@ class DistEmbedding:
 
     def __call__(self, idx, device=th.device("cpu")):
         """
-        idx : th.tensor
+        node_ids : th.tensor
             Index of the embeddings to collect.
         device : th.device
             Target device to put the collected embeddings.
@@ -109,6 +109,7 @@ class DistEmbedding:
         Tensor
             The requested node embeddings
         """
+        idx = utils.toindex(idx).tousertensor()
         emb = self._tensor[idx].to(device, non_blocking=True)
         if F.is_recording():
             emb = F.attach_grad(emb)

python/dgl/distributed/utils.py

-""" Utility functions for distributed training."""
-
-import torch
-
-from ..utils import toindex
-
-
-def totensor(data):
-    """Convert the given data to a tensor.
-
-    Parameters
-    ----------
-    data : tensor, array, list or slice
-        Data to be converted.
-
-    Returns
-    -------
-    Tensor
-        Converted tensor.
-    """
-    if isinstance(data, torch.Tensor):
-        return data
-    return toindex(data).tousertensor()

tests/distributed/test_dist_tensor.py

@@ -57,7 +57,12 @@ def check_binary_op(key1, key2, key3, op):
             dist_g.edata[key3][i:i_end],
             op(dist_g.edata[key1][i:i_end], dist_g.edata[key2][i:i_end]),
         )
-        _ = dist_g.edata[key3][F.tensor([100, 20, 10], F.int32)]
+        # Test with different index dtypes. int32 is not supported.
+        with pytest.raises(
+            dgl.utils.internal.InconsistentDtypeException,
+            match="DGL now requires the input tensor to have",
+        ):
+            _ = dist_g.edata[key3][F.tensor([100, 20, 10], F.int32)]
         _ = dist_g.edata[key3][F.tensor([100, 20, 10], F.int64)]