SizePartitioner¶

class SizePartitioner(partition_sizes: Sequence[int])[source]¶

Bases: Partitioner

Partitioner that creates each partition with the size specified by a user.

Parameters:: partition_sizes (Sequence[int]) – The size of each partition. partition_id 0 will have partition_sizes[0] samples, partition_id 1 will have partition_sizes[1] samples, etc.

Examples

>>> from flwr_datasets import FederatedDataset
>>> from flwr_datasets.partitioner import SizePartitioner
>>>
>>> partition_sizes = [15_000, 5_000, 30_000]
>>> partitioner = SizePartitioner(partition_sizes)
>>> fds = FederatedDataset(dataset="cifar10", partitioners={"train": partitioner})

Methods

`is_dataset_assigned`()	Check if a dataset has been assigned to the partitioner.
`load_partition`(partition_id)	Load a single partition of the size of partition_sizes[partition_id].

Attributes

`dataset`	Dataset property.
`num_partitions`	Total number of partitions.
`partition_id_to_indices`	Partition id to indices (the result of partitioning).

property dataset: Dataset¶: Dataset property.

is_dataset_assigned() → bool¶

Check if a dataset has been assigned to the partitioner.

This method returns True if a dataset is already set for the partitioner, otherwise, it returns False.

Returns:: dataset_assigned – True if a dataset is assigned, otherwise False.
Return type:: bool

load_partition(partition_id: int) → Dataset[source]¶

Load a single partition of the size of partition_sizes[partition_id].

For example if given partition_sizes=[20_000, 10_000, 30_000], then partition_id=0 will return a partition of size 20_000, partition_id=1 will return a partition of size 10_000, etc.

Parameters:: partition_id (int) – The index that corresponds to the requested partition.
Returns:: dataset_partition – Single dataset partition.
Return type:: Dataset

property num_partitions: int¶: Total number of partitions.

property partition_id_to_indices: dict[int, list[int]]¶: Partition id to indices (the result of partitioning).