osl_dynamics.data.tf.dataset_from_index#

osl_dynamics.data.tf.dataset_from_index(index, slice_fn, batch_size, shuffle=True, drop_last_batch=False, seed=None)#

Create a batched dataset from a sequence index.

Shuffling is done on the index (which is tiny), so it is a perfect shuffle without a buffer. The index is reshuffled every epoch.

Parameters:
indexnp.ndarray

Sequence index from sequence_index().

slice_fncallable

Function from make_slice_fn().

batch_sizeint

Number sequences in each mini-batch which is used to train the model.

shufflebool, optional

Should we shuffle the sequences?

drop_last_batchbool, optional

Should we drop the last batch if it is smaller than the batch size?

seedint, optional

Seed for shuffling.

Returns:
datasettf.data.Dataset

Batched dataset. The number of batches is known (dataset.cardinality()).