SingleSessionDataset¶

class behavenet.data.data_generator.SingleSessionDataset(data_dir, lab='', expt='', animal='', session='', signals=None, transforms=None, paths=None, device='cuda', as_numpy=False)[source]¶

Bases: SingleSessionDatasetBatchedLoad

Dataset class for a single session.

Loads all data during Dataset creation and saves as an attribute. Batches are then sampled from this stored data. All data transformations are applied to the full dataset upon load, not for each batch. This automatically returns data as lists of numpy arrays.

Note

This data loader cannot be used to fit pytorch models, only ssm models.