add warning

torch_cluster/rw.py

+import warnings
+
 import torch
 
 if torch.cuda.is_available():
     import torch_cluster.rw_cuda
 
 
-def random_walk(row, col, start, walk_length, num_nodes=None):
+def random_walk(row, col, start, walk_length, p=1, q=1, num_nodes=None):
+    if p != 1 or q != 1:
+        warnings.warn('Parameters `p` and `q` are not supported yet and will'
+                      'be restored to their default values `p=1` and `q=1`.')
+        p = q = 1
+
     num_nodes = row.max().item() + 1 if num_nodes is None else num_nodes
     if row.is_cuda:
-        return torch_cluster.rw_cuda.rw(row, col, start, walk_length, 1, 1,
+        return torch_cluster.rw_cuda.rw(row, col, start, walk_length, p, q,
                                         num_nodes)
     else:
         raise NotImplementedError