autointent.context.data_handler.check_split_readiness

autointent.context.data_handler.check_split_readiness(dataset, split, config, allow_oos_in_train=None)

Check whether the dataset has enough samples per class for autointent pipeline.

Parameters:

• dataset (autointent.Dataset) – The dataset to check (e.g. the same passed to split_dataset()).

• split (str) – The split name to check (e.g. Split.TRAIN).

• config (autointent.configs.DataConfig) – data config

• allow_oos_in_train (bool | None) – Same as in split_dataset(). If the split contains OOS samples and this is None, this function raises ValueError (mirrors splitting behavior).

Return type:

SplitReadinessResult