autointent.context.data_handler.SplitReadinessResult#
- class autointent.context.data_handler.SplitReadinessResult#
Result of checking whether a dataset can be fed to autointent pipeline.
- ready#
True if stratification can be performed (enough samples per class).
- underpopulated_classes#
List of (label, n_samples) for classes below the minimum.
- min_samples_per_class_required#
Minimum samples per class used for the check.
- reason#
Human-readable reason when not ready (e.g. OOS not configured).