v0.7.8

MLDatasets v0.7.8

Diff since v0.7.7

Closed issues:

• Some TU Datasets do not work - Error when loading (#195)
• Potential edge mismatch in TUDatasets? (#197)

Merged pull requests:

• Support multiple labels for nodes, edges and graphs (#196) (@Dsantra92)
• [tudataset] wrong perm (#198) (@CarloLucibello)

v0.7.7

MLDatasets v0.7.7

Diff since v0.7.6

Closed issues:

• Make an unifying type for all graph datasets (#169)
• bug in MovieLens indices (#190)

Merged pull requests:

• Fix typos (#187) (@aurorarossi)
• fix PolBlogs (#188) (@CarloLucibello)
• fix UD_English (#189) (@CarloLucibello)
• don't add reverse edges in movielens (#193) (@CarloLucibello)

v0.7.6

MLDatasets v0.7.6

Diff since v0.7.5

Closed issues:

• load torch tensors in OGBDatasets (#107)
• World age issue wtih read_pytorch (#167)

Merged pull requests:

• require explicit Pickle import in load_pytorch (#170) (@Dsantra92)
• Split for OGBDatasets (#172) (@Dsantra92)
• add OrganicMaterialsDB (#180) (@CarloLucibello)
• Corrects docstring for the as_df keyword (#183) (@vitaminace33)
• CompatHelper: bump compat for MLUtils to 0.3, (keep existing compat) (#186) (@github-actions[bot])

v0.7.5

MLDatasets v0.7.5

Diff since v0.7.4

Closed issues:

• MPI Faust dataset (#155)
• Issues on M1 due to use of BinaryProvider (#159)
• Add test cases for PTBLM dataset (#163)
• Graph examples are not working (#176)
• Module are not callable (#179)

Merged pull requests:

• Add Faust Dataset (#160) (@Dsantra92)
• Fix the references in MovieLens (#161) (@Dsantra92)
• Don't test SMS Spam Collection on CI (#164) (@Dsantra92)
• add PTBLM tests (#165) (@Dsantra92)

v0.7.4

MLDatasets v0.7.4

Diff since v0.7.3

Closed issues:

• New test failing ("ml-10m") (#151)

Merged pull requests:

• Fix documentation typos for vision datasets (#150) (@christiangnrd)
• Add Omniglot Dataset (#152) (@christiangnrd)
• Patch/fix ci (#154) (@Dsantra92)
• Fix movielens 10m (#156) (@christiangnrd)
• Reorder tests and documentation lists in alphabetical order for consistency (#157) (@christiangnrd)

v0.7.3

MLDatasets v0.7.3

Diff since v0.7.2

Closed issues:

• Movielens datasets (#104)
• Support for Heterogeneous Graphs (#139)
• Don't download non-required data for DataHubs (#142)

Merged pull requests:

• Use Manual deps for registering some datahubs (#143) (@Dsantra92)
• MovieLens Datasets (#147) (@Dsantra92)
• fixed CIFAR100 docs class number error (#149) (@cluffa)

v0.7.2

MLDatasets v0.7.2

Diff since v0.7.1

Closed issues:

• Contributing synthetic datasets (#29)
• read_csv needs DataFrames.jl which is a require import (#133)
• add show methods for WithGraph and GNNChain (#134)
• Possible World Age issue when require error is not invoked (#135)
• ] add MLDatasets fails on Julia nightly on macOS (#137)

Merged pull requests:

• eagerly load CSV (#136) (@johnnychen94)
• Support for OGB Hetero-Graphs (#140) (@Dsantra92)
• Fix unzip errors in OGB-Datasets (#141) (@Dsantra92)

v0.7.1

MLDatasets v0.7.1

Diff since v0.7.0

Merged pull requests:

• KarateClub <: AbstractDataset (#132) (@CarloLucibello)

v0.7.0

MLDatasets v0.7.0

Diff since v0.6.0

Closed issues:

• Add documentation pages for NLP datasets (#14)
• Problem running Pkg.test (#16)
• fail to compile on Julia 1.1 Linux (#27)
• redesign the package (#73)
• Links to other dataset repositories (#115)
• using MLDatasets is very slow (#126)
• missing docstring pyhasattr in docs (#129)

Merged pull requests:

• lazy module loading (#128) (@CarloLucibello)

v0.6.0

MLDatasets v0.6.0

Diff since v0.5.16

Closed issues:

• Deprecations v0.6 (#120)

Merged pull requests:

• add SupervisedDataset and move misc. datasets to the new interface (#98) (@CarloLucibello)
• Redesign Vision datasets (#110) (@CarloLucibello)
• fix docs (#111) (@CarloLucibello)
• Add Zachary's Karate Club Dataset (#112) (@Dsantra92)
• Redesign Text datasets (#113) (@CarloLucibello)
• Redesign graph datasets (#114) (@CarloLucibello)
• Reddit Graph dataset (#116) (@Dsantra92)
• typo: README url link (#117) (@athatheo)
• redesign TUDataset (#119) (@CarloLucibello)
• Add missing deprecations for v0.6 (#121) (@CarloLucibello)
• docs improvements (#123) (@CarloLucibello)
• change dataset[] to dataset[:] (#124) (@CarloLucibello)