src.cross_validate

Module Contents

Classes

CrossValidation

class src.cross_validate.CrossValidation(saving_dir: str, n_splits: int = 5)

device

fit(model: src.solutions.base_solution.BaseSolution, X: pandas.DataFrame, y: pandas.DataFrame) → pandas.DataFrame

Makes average fold prediction

Parameters

• model – predictor from BaseSolution class

• X – Dataframe that has text_id and full_text columns

• y – Dataframe that has text_id, cohesion, … columns

Returns

Dataframe with class scores for each split and overall CV score

predict(X: pandas.DataFrame) → pandas.DataFrame

Makes average fold prediction

Parameters

X – Dataframe that have text_id and full_text columns

Returns

prediction Dataframe that have text_id, cohesion, … columns

save(path: Union[str, pathlib.Path])

load(path: Union[str, pathlib.Path], predictor: src.solutions.base_solution.BaseSolution)