default_suitesΒΆ

Functions for loading the default (built-in) suites for various validation stages.

Each function returns a new suite that is initialized with a list of checks and default conditions. It is possible to customize these suites by editing the checks and conditions inside it after suite creation.

Functions

single_dataset_integrity()

Create a suite that is meant to detect integrity issues within a single dataset.

train_test_leakage()

Create a suite that is meant to detect data leakage between the training dataset and the test dataset.

train_test_validation()

Create a suite that is meant to validate correctness of train-test split, including integrity, distribution and leakage checks.

model_evaluation()

Create a suite that is meant to test model performance and overfit.

full_suite()

Create a suite that includes many of the implemented checks, for a quick overview of your model and data.

previous

Suites

next

single_dataset_integrity