[Example] Temporal Graph Neural Network #2636
Merged
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
sneakerkg
requested changes
Feb 19, 2021
sneakerkg
reviewed
Feb 26, 2021
examples/pytorch/tgn/data.py
Outdated
| new_df.i += 1 | ||
| new_df.idx += 1 | ||
|
|
||
| return new_df |
There was a problem hiding this comment.
If the code is adapted from original code, please add a commet: https://github.com/dmlc/dgl/blob/master/examples/pytorch/pointcloud/pointnet/pointnet2.py#L10
examples/pytorch/tgn/data.py
Outdated
| train_div = int(0.7*num_edges) | ||
| valid_div = int(0.85*num_edges) | ||
|
|
||
| self.nn_test_g = dgl.edge_subgraph(self.g,range(valid_div,num_edges)) |
examples/pytorch/tgn/data.py
Outdated
| edge_ids = range(self.batch_cnt*self.batch_size,min((self.batch_cnt+1)*self.batch_size,graph.num_edges())) | ||
| subgraph = dgl.edge_subgraph(graph,edge_ids) | ||
| working_d = DictNode(parent=p_dict,NIDdict=subgraph.ndata[dgl.NID]) | ||
| subgraph.ndata[dgl.NID] = torch.from_numpy(working_d.GetRootID(range(subgraph.num_nodes()))) |
There was a problem hiding this comment.
Don't use dgl.NID. Currently the behavior of dgl.NID is not clearly defined and this is a dgl system setting. After we get clear the behavior of dgl.NID in system level, we may revisit here.
examples/pytorch/tgn/data.py
Outdated
| map_index = self.NIDdict[index] | ||
| return self.parent.GetRootID(map_index) | ||
|
|
||
| class TemporalDataLoader: |
There was a problem hiding this comment.
Try to inherit from pytorch dataloader.
examples/pytorch/tgn/tgn.py
Outdated
|
|
||
| # Fake link embedding and compute score | ||
| new_neg,neg_subg = self.node_sampler(neg,ts,mode) | ||
| n_ts = ts.repeat(neg_subg.num_nodes()) |
examples/pytorch/tgn/tgn.py
Outdated
| pred_pos = self.linkpredictor(embedding[pos_src_id],embedding[pos_dst_id]) | ||
| pred_neg = self.linkpredictor(embedding[pos_src_id],embedding[neg_id]) | ||
| subg = dgl.remove_self_loop(subg) | ||
| subg = dgl.add_self_loop(subg) |
Add a simple temporal graph sampling method. This method can support k-hop sampling while others can not. This method also achieve good performance and speed.
sneakerkg
requested changes
Mar 22, 2021
sneakerkg
approved these changes
Mar 22, 2021
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Description
Checklist
Please feel free to remove inapplicable items for your PR.
or have been fixed to be compatible with this change
Changes