rnaglib.dataset_transforms.DefaultBasePairLoader

class rnaglib.dataset_transforms.DefaultBasePairLoader(dataset=None, data_path=None, batch_size=5, inner_batch_size=50, sampler_layers=2, neg_samples=1, num_workers=4, framework='dgl', **kwargs)

Dataloader that yields base pairs.

Just a default edge base pair loader that deals with the splits.

Parameters:

• dataset – A GraphDataset we want to loop over for base-pair prediction

• data_path – Optionally, we can use a data path to create a default GraphDataset

• batch_size – The desired batch size (number of whole graphs)

• inner_batch_size – The desired inner batch size (number of sampled edge in a batched graph)

• sampler_layers – The size of the neighborhood

• neg_samples – The number of negative sample to use per positive ones

• framework – Which learning framework to use

• num_workers – The number of cores to use for loading

__init__(dataset=None, data_path=None, batch_size=5, inner_batch_size=50, sampler_layers=2, neg_samples=1, num_workers=4, framework='dgl', **kwargs)

Just a default edge base pair loader that deals with the splits.

Parameters:

• dataset – A GraphDataset we want to loop over for base-pair prediction

• data_path – Optionally, we can use a data path to create a default GraphDataset

• batch_size – The desired batch size (number of whole graphs)

• inner_batch_size – The desired inner batch size (number of sampled edge in a batched graph)

• sampler_layers – The size of the neighborhood

• neg_samples – The number of negative sample to use per positive ones

• framework – Which learning framework to use

• num_workers – The number of cores to use for loading

Methods

__init__([dataset, data_path, batch_size, ...])	Just a default edge base pair loader that deals with the splits.
get_data()