Lightrun Answers was designed to reduce the constant googling that comes with debugging 3rd party libraries. It collects links to all the places you might be looking at while hunting down a tough bug.
And, if you’re still stuck at the end, we’re happy to hop on a call to see how we can help out.
k_hop_subgraph cannot maintain that the actual mini-batch nodes always be placed first
See original GitHub issue🛠 Proposed Refactor
I am glad to find that actual mini-batch nodes always be placed first
https://github.com/pyg-team/pytorch_geometric/discussions/5279#discussioncomment-3814170
But
https://github.com/pyg-team/pytorch_geometric/blob/2d77047811311544d4861fa9e80fd2ec60ab6d01/torch_geometric/utils/subgraph.py#L302
unique always always sort the tensor link
So that it cannot maintain that the actual mini-batch nodes always be placed first.
Suggest a potential alternative/fix
I just use set.
origin_nodes = set([node_idx])
for i in range(1,num_hops):
subsets[i] = subsets[i][~np.isin(subsets[i],origin_nodes)]
But I don’t know how to return_inverse correctly.
Issue Analytics
- State:
- Created a year ago
- Comments:8 (8 by maintainers)
Top Results From Across the Web
Top Related Medium Post
Top Related StackOverflow Question
Troubleshoot Live Code
Top Related Reddit Thread
Top Related Hackernoon Post
Top Related Tweet
Top Related Dev.to Post
Top Related Hashnode Post
@LukeLIN-web The implementation is here.
At best, we use the implementation from
pyg-libas used in theNeighborLoader, see here. It is documented here.