| from src.cross_validate import CrossValidation | |
| from src.data_reader import load_train_test_df | |
| from src.solutions.constant_predictor import ConstantPredictorSolution | |
| from src.utils import pandas_set_print_options | |
| pandas_set_print_options() | |
| def test_cross_validation(): | |
| train_df, _ = load_train_test_df(is_testing=True) | |
| x_columns = ['text_id', 'full_text'] | |
| X, y = train_df[x_columns], train_df.drop(columns=['full_text']) | |
| n_splits = 3 | |
| cv = CrossValidation(saving_dir='/tmp/sdfjsld', n_splits=n_splits) | |
| predictor = ConstantPredictorSolution() | |
| cv_scores = cv.fit(predictor, X, y) | |
| assert cv_scores.shape == (n_splits + 1, 6) | |
| prediction_df = cv.predict(train_df[x_columns]) | |
| assert prediction_df.shape == (len(train_df), 7) | |