v0.6.3

GraphNeuralNetworks v0.6.3

Diff since v0.6.2

Closed issues:

• Julia Formatter (#238)
• Update documentation: Convolutional Layers (#255)

Merged pull requests:

• add JuliaFormatter to Github Workflows (#253) (@pri1311)
• fix conv table (#256) (@CarloLucibello)
• CompatHelper: bump compat for CUDA to 4, (keep existing compat) (#257) (@github-actions[bot])

v0.6.2

GraphNeuralNetworks v0.6.2

Diff since v0.6.1

Closed issues:

• A Logo is needed (#35)

Merged pull requests:

• faster unbatch (#248) (@CarloLucibello)
• Add a transformer-like convolutional layer (#249) (@graidl)
• Add new cache for changed notebooks! (#250) (@Dsantra92)
• CompatHelper: bump compat for MLUtils to 0.4, (keep existing compat) (#252) (@github-actions[bot])

v0.6.1

GraphNeuralNetworks v0.6.1

Diff since v0.6.0

Closed issues:

• Overflows in GATConv and GATv2Conv (#246)

Merged pull requests:

• fixing overflow in GAT layers (#244) (@pevnak)
• numerically stable GATv2Conv (#247) (@CarloLucibello)

v0.6.0

GraphNeuralNetworks v0.6.0

Diff since v0.5.2

Closed issues:

• Heterogeneous graphs support (#144)
• Support for Heterogeneous Graphs (#199)
• Remove hard dependency on GraphNeuralNetworks from Pluto Notebooks (#204)
• Generation of documentation is very slow because of Pluto (#227)
• GNN.jl in tutorials (#228)
• GNNGraph with multiple edge features not working (#243)

Merged pull requests:

• HeteroGraph implementation (#146) (@CarloLucibello)
• Fix typo in docstring (#226) (@aurorarossi)
• Remove GNN.jl (#229) (@aurorarossi)
• Cache Pluto Notebooks (#230) (@Dsantra92)
• convenience accessors (#231) (@CarloLucibello)
• implement DataStore (#232) (@CarloLucibello)
• CompatHelper: bump compat for Functors to 0.4, (keep existing compat) (#233) (@github-actions[bot])
• Compact print of GNNGraphs (#234) (@aurorarossi)
• Fix small typos (#235) (@aurorarossi)
• Improve show function (#236) (@aurorarossi)
• Improve hetero show function (#237) (@aurorarossi)
• Implemented khop_adj (#239) (@aurorarossi)
• Implement laplacian_lambda_max (#240) (@aurorarossi)
• Fix gnnheterographs failed test (#241) (@aurorarossi)
• update tutorials (#245) (@CarloLucibello)

v0.5.2

GraphNeuralNetworks v0.5.2

Diff since v0.5.1

Closed issues:

• NNConv tests are failing (#208)

Merged pull requests:

• Avoid capturing external variables into closures (#225) (@CarloLucibello)

v0.5.1

GraphNeuralNetworks v0.5.1

Diff since v0.5.0

Closed issues:

• Add tutorials written in Pluto (#168)
• copy(::GNNGraph)? (#191)
• Allow for additonal features in GNNGraph (#210)
• support equivariant neural networks (#221)

Merged pull requests:

• implement copy (#219) (@CarloLucibello)
• equivariant gnn (#223) (@CarloLucibello)
• doc fixes (#224) (@CarloLucibello)

v0.5.0

GraphNeuralNetworks v0.5.0

Diff since v0.4.5

Closed issues:

• don't automatically batch when getobs from an array of graphs (#170)
• Question about the GCNConv layer code (#211)
• Support for multiple graphs in GNNGraph (#212)

Merged pull requests:

• don't automatically batch graphs inside vector when using getobs (#183) (@CarloLucibello)
• Node classification tutorial (#198) (@Dsantra92)
• Fixed a few typos in notebooks, docs and docstrings (#205) (@pitmonticone)
• Switch to DemoCards.jl (#206) (@Dsantra92)
• temporarily remove NNConv tests (#209) (@CarloLucibello)
• Switch to default pluto frontmatter (#214) (@Dsantra92)
• CompatHelper: bump compat for KrylovKit to 0.6, (keep existing compat) (#216) (@github-actions[bot])
• CompatHelper: bump compat for MLUtils to 0.3, (keep existing compat) (#217) (@github-actions[bot])

v0.4.5

GraphNeuralNetworks v0.4.5

Diff since v0.4.4

Closed issues:

• Construct Graphs.SimpleDiGraph graphs from GNNGraphs (#167)
• add radius_graph api (#177)
• knn_graph yields same results with or without self loops (#179)
• Missing bounds checking when working on GPU (#181)
• Spelling Error of edge? (#184)
• Duplicate indexes in documents (#187)
• Implement GraphWorld for fake graphs benchmarking (#188)
• Some design issues (#193)
• Flux destructure/restructure bug (#200)

Merged pull requests:

• add radius_graph api (#180) (@MilkshakeForReal)
• size checks in apply_edges and aggregate_neighbors (#182) (@CarloLucibello)
• Fix Spelling (#186) (@umbriquse)
• don't display redundant docstrings from Graphs.jl (#192) (@CarloLucibello)
• CompatHelper: bump compat for Functors to 0.3, (keep existing compat) (#194) (@github-actions[bot])
• Update conv.md (#195) (@pitmonticone)
• [Test] Render pluto files in the CI (#196) (@Dsantra92)
• Fix typo (#197) (@MilkshakeForReal)
• fix GNNChain restructure bug (#201) (@mplemay)
• Update GNNChain (#202) (@CarloLucibello)
• Replace depreciated Flux.ADAM with Flux.Adam (#203) (@mplemay)

v0.4.4

GraphNeuralNetworks v0.4.4

Diff since v0.4.3

Closed issues:

• Inaccurate GATv2Conv Documentation (#175)

Merged pull requests:

• fixing CI for pluto tutorials (#174) (@CarloLucibello)
• Documentation fixes for GATv2Conv (#176) (@Animiral)

v0.4.3

GraphNeuralNetworks v0.4.3

Diff since v0.4.2

Merged pull requests:

• add Pluto tutorials (#172) (@CarloLucibello)