morphoclass.data.morphology_data_loader module

Morphology data loader.

class morphoclass.data.morphology_data_loader.MorphologyDataLoader(dataset: morphoclass.data.morphology_dataset.MorphologyDataset, **kwargs: Any)

Bases: Generic[torch.utils.data.dataloader.T_co]

A data loader for the morphology data set.

This class is derived from torch.utils.data.DataLoader and unlike torch_geometric.data.DataLoader is able to handle Data objects with non-numeric fields. These fields are simply ignored upon constructing batches.

Parameters

• dataset (torch.utils.data.dataset.Dataset[T_co]) – The data set to apply the data loader to.

• kwargs – Further parameter to pass on to the PyTorch DataLoader base class.

batch_size: Optional[int]

dataset: torch.utils.data.dataset.Dataset[T_co]

drop_last: bool

num_workers: int

pin_memory: bool

prefetch_factor: int

sampler: Union[torch.utils.data.sampler.Sampler, Iterable]

timeout: float