API#

Loader#

Loader(*[, batch_sampler, chunk_size, ...])

A loader for on-disk data anndata stores.

Loader.__iter__()

Iterate over the on-disk datasets.

io helpers#

write_sharded(group, adata, *[, ...])

Write a sharded zarr store from a single AnnData object.

DatasetCollection(group, *[, mode, ...])

A preshuffled collection object including functionality for creating, adding to, and loading collections shuffled by annbatch.

Samplers#

ClassSampler(chunk_size, preload_nchunks, ...)

Sample class-coherent batches with replacement.

RandomSampler(chunk_size, preload_nchunks, ...)

Shuffled chunk-based sampler for batched data access.

SequentialSampler(chunk_size, ...[, ...])

Ordered chunk-based sampler for batched data access.

DistributedSampler(sampler, *, dist_info[, ...])

Distributed chunk-based sampler that shards data across distributed processes.

abc#

Sampler()

Base sampler class.

types#

LoaderOutput

The output of the loader, the "data matrix" with its obs, optional, var, optional, and index, also optional.

LoadRequest

Load request from sampler.