DataSplits#

class DataSplits(n_splits: int)[source]#

Bases: object

Container for DataSplitInfo instances created from the same dataset.

This class provides a fixed-capacity container for managing multiple DataSplitInfo instances that represent different splits of the same dataset. It ensures type safety and provides index-based access to individual splits.

Parameters:

n_splitsint: The number of splits to be stored in the container. Must be a positive integer.

Attributes:

_data_splitsList[DataSplitInfo or None]: A list of DataSplitInfo instances, initialized with None values
_current_indexint: The index where the next split will be added
expected_n_splitsint: The total number of splits the container can hold

Raises:

ValueError: If n_splits is not a positive integer

Notes

The container is initialized with a fixed capacity and fills slots sequentially. Once all slots are filled, no more splits can be added.

Examples

Create a container for 5 splits:

>>> splits = DataSplits(n_splits=5)
>>> print(len(splits))  # 5

Add splits to the container:

>>> for i in range(5):
...     split_info = DataSplitInfo(...)  # Create split
...     splits.add(split_info)

Access a specific split:

>>> split_0 = splits.get_split(0)
>>> split_2 = splits.get_split(2)

add(split: DataSplitInfo) → None[source]#

Add a DataSplitInfo instance to the container.

Adds a DataSplitInfo instance to the next available slot in the container. The split is added sequentially, starting from index 0.

Parameters:

splitDataSplitInfo: The DataSplitInfo instance to add to the container

Raises:

IndexError: If the number of splits exceeds the expected number of splits
TypeError: If the split is not a DataSplitInfo instance

Notes

The method adds splits sequentially, incrementing the current index after each addition. Once all slots are filled, no more splits can be added.

Examples

Add a split to the container:

>>> splits = DataSplits(n_splits=3)
>>> split_info = DataSplitInfo(...)
>>> splits.add(split_info)

Attempt to add more splits than capacity:

>>> for i in range(4):  # More than n_splits=3
...     splits.add(DataSplitInfo(...))  # Raises IndexError

get_split(index: int) → DataSplitInfo[source]#

Get a DataSplitInfo instance by index.

Retrieves the DataSplitInfo instance at the specified index. Validates that the index is within bounds and that a split exists at that index.

Parameters:

indexint: The index of the DataSplitInfo instance to retrieve. Must be between 0 and expected_n_splits - 1

Returns:

DataSplitInfo: The DataSplitInfo instance at the specified index

Raises:

IndexError: If the index is out of bounds (not between 0 and expected_n_splits - 1)
ValueError: If no DataSplitInfo instance has been assigned to the specified index

Notes

The method performs two validation checks: 1. Ensures the index is within the valid range 2. Ensures a split has been assigned to that index

Examples

Get a split by index:

>>> splits = DataSplits(n_splits=3)
>>> splits.add(DataSplitInfo(...))  # Adds to index 0
>>> split_0 = splits.get_split(0)

Attempt to access out-of-bounds index:

>>> splits.get_split(5)  # Raises IndexError

Attempt to access unassigned index:

>>> splits.get_split(1)  # Raises ValueError (no split at index 1)

DataSplits#

This Page