bluecast.tests.test_experiment_tracker

Module Contents

Functions

experiment_tracker()

Create a temporary DuckDB-based experiment tracker for testing.

sample_training_config()

Create a sample training config for testing.

test_add_hyperparameter_results(experiment_tracker, ...)

Test adding hyperparameter tuning results.

test_add_evaluation_results(experiment_tracker, ...)

Test adding evaluation results.

test_retrieve_results_as_df(experiment_tracker, ...)

Test retrieving results as DataFrame.

test_get_best_score_empty(experiment_tracker)

Test getting best score when no results exist.

test_get_best_score_higher_is_better(...)

Test getting best score for higher-is-better metric.

test_get_best_score_lower_is_better(...)

Test getting best score for lower-is-better metric.

test_get_experiment_summary(experiment_tracker, ...)

Test getting experiment summary statistics.

test_legacy_compatibility(experiment_tracker, ...)

Test legacy compatibility properties.

test_database_persistence()

Test that data persists across tracker instances.

test_temporary_database()

Test that temporary database is created when no path provided.

bluecast.tests.test_experiment_tracker.experiment_tracker()

Create a temporary DuckDB-based experiment tracker for testing.

bluecast.tests.test_experiment_tracker.sample_training_config()

Create a sample training config for testing.

bluecast.tests.test_experiment_tracker.test_add_hyperparameter_results(experiment_tracker, sample_training_config)

Test adding hyperparameter tuning results.

bluecast.tests.test_experiment_tracker.test_add_evaluation_results(experiment_tracker, sample_training_config)

Test adding evaluation results.

bluecast.tests.test_experiment_tracker.test_retrieve_results_as_df(experiment_tracker, sample_training_config)

Test retrieving results as DataFrame.

bluecast.tests.test_experiment_tracker.test_get_best_score_empty(experiment_tracker)

Test getting best score when no results exist.

bluecast.tests.test_experiment_tracker.test_get_best_score_higher_is_better(experiment_tracker, sample_training_config)

Test getting best score for higher-is-better metric.

bluecast.tests.test_experiment_tracker.test_get_best_score_lower_is_better(experiment_tracker, sample_training_config)

Test getting best score for lower-is-better metric.

bluecast.tests.test_experiment_tracker.test_get_experiment_summary(experiment_tracker, sample_training_config)

Test getting experiment summary statistics.

bluecast.tests.test_experiment_tracker.test_legacy_compatibility(experiment_tracker, sample_training_config)

Test legacy compatibility properties.

bluecast.tests.test_experiment_tracker.test_database_persistence()

Test that data persists across tracker instances.

bluecast.tests.test_experiment_tracker.test_temporary_database()

Test that temporary database is created when no path provided.