AutoDataset

class hivegraph.io.autodataset.AutoDataset(dataset_name: str, path: str = 'artifacts/data/', sparse: bool = True, cleaned: bool = False)[source]

Bases: object

Helper class to instantiate a dataset based on the dataset name.

Parameters:

dataset_name (str) – Name of the dataset.
path (str, optional) – Path to the dataset. Defaults to “artifacts/data/”.
sparse (bool, optional) – If True, the dataset is returned as a sparse dataset. Defaults to True.
cleaned (bool, optional) – If True, the dataset is returned as a cleaned dataset. Defaults to False.

References:

https://github.com/pyg-team/pytorch_geometric/blob/master/benchmark/kernel/datasets.py

Raises:: AssertionError – If the dataset name is not supported.

get_dataset() → Dataset[source]

Fetches the dataset

Returns:: PyG dataset.
Return type:: Dataset

download() → Dataset[source]

Downloads the dataset

Returns:: PyG dataset.
Return type:: Dataset
Raises:: NotImplementedError – If the dataset is not implemented.

process() → Dataset[source]

Processes the dataset

Returns:: PyG dataset.
Return type:: Dataset
Raises:: NotImplementedError – If the dataset is not implemented.

process_classification_dataset() → Dataset[source]

Processes the classification dataset

Returns:: PyG dataset.
Return type:: Dataset

process_transductive_dataset() → Dataset[source]

Processes the transductive dataset

Returns:: PyG dataset.
Return type:: Dataset