cleanup

Author: CarloLucibello

Project.toml

@@ -1,7 +1,7 @@
 name = "PyGDatasets"
 uuid = "44f67abd-f36e-4be4-bfe5-65f468a62b3d"
 authors = ["Carlo Lucibello <carlo.lucibello@gmail.com> and contributors"]
-version = "0.1.1"
+version = "0.1.2"
 
 [deps]
 DLPack = "53c2dc0f-f7d5-43fd-8906-6c0220547083"

README.md

@@ -52,7 +52,7 @@ load_dataset("MoleculeNet", name="ESOL")
 
 # heterographs
 load_dataset("IMDB")
-load_dataset("AMiner")
+load_dataset("MovieLens100K")
 ```
 
 ## Dataset References

src/load_dataset.jl

@@ -7,7 +7,7 @@ and return it as an  [`InMemoryGNNDataset`](@ref) object.
 The keyword arguments `kws` are passed to the dataset constructor.
 
 See [`pygdata_to_gnngraph`](@ref) for details on the conversion of 
-PyTorch Geometric `Data` objects to `GNNGraph`s.
+PyTorch Geometric `Data` and `HeteroData` objects to `GNNGraph` and `GNNHeteroGraph` objects.
 
 For PyG datasets requiring a `root` argument, the default value 
 `"\$(PyGDatasets.DEFAULT_ROOT[])/\$name"` is provided if not specified.
@@ -21,11 +21,9 @@ This scratch space will be deleted when the package is removed.
 # Examples
 
 ```julia
-load_dataset("TUDataset", root="./", name="MUTAG")
-# is equivalent to the python code
-# pyg.datasets.TUDataset(root="./", name="MUTAG")
-
-load_dataset("Planetoid", name="Cora")
+dataset = load_dataset("TUDataset", root="./", name="MUTAG")
+dataset = load_dataset("Planetoid", name="Cora") # use default root
+dataset = load_dataset("IMDB") # heterogenous graph dataset
 ```
 """
 function load_dataset(dataset::String; root=nothing, kws...)