Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We鈥檒l occasionally send you account related emails.

Already on GitHub? Sign in to your account

[RFC] 0.5 Sampler Design for Heterographs #1199

Open
BarclayII opened this issue Jan 13, 2020 · 1 comment
Open

[RFC] 0.5 Sampler Design for Heterographs #1199

BarclayII opened this issue Jan 13, 2020 · 1 comment

Comments

@BarclayII
Copy link
Collaborator

@BarclayII BarclayII commented Jan 13, 2020

馃殌 Feature

We have finally settled down (in principle) the design of the long-awaited sampling on heterogeneous graphs, and we have started working on it.

This issue keeps track of the implementation progress.

You're welcome to read the proposal publicly available on https://docs.google.com/document/d/1cVvbM_QtaIN0UsqrgB0FhlYti1PjXua2QbSAoMaE2ng/edit

The document is not fixed and it is subject to changes: you're welcome to provide feedback on either the proposal itself or this issue thread.

Task tracker

  • C APIs
    • Neighbor sampling [@jermainewang ]
      • sample_neighbors
      • sample_neighbors_topk
      • in_subgraph
      • out_subgraph
      • sample_layerwise
    • Random walks
    • Misc
      • dgl::aten refactor to cover float types as well (only necessary ones)
      • create_graph_from_paths [@BarclayII ]
      • compact_graphs [@BarclayII ]
      • to_simple_graph [@BarclayII ]
      • dgl.choice(prob_array, num_samples) (faster numpy.random.choice) [@yzh119, already done in #1142 , may a python interface is required? ]
      • Graph.remove_edges() [@BarclayII ]
  • Models (rewrite)

Related:

  • Unification of DGLGraph, DGLHeteroGraph, SubGraph, BatchedGraph, etc.
  • Shared memory storage
@BarclayII

This comment has been minimized.

Copy link
Collaborator Author

@BarclayII BarclayII commented Jan 13, 2020

@yzh119 I'm thinking of putting the dgl.choice as RandomEngine::ThreadLocal()->Choice(FloatArray unnormalized_prob)

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Projects
None yet
Linked pull requests

Successfully merging a pull request may close this issue.

None yet
1 participant
You can鈥檛 perform that action at this time.