catboost_regressor.py

"""
CatBoost Regressor Example.

This script demonstrates the usage of the CatBoost Regressor for a simple regression task.
CatBoost is a powerful gradient boosting library that handles categorical features
automatically and is highly efficient.

Make sure to install CatBoost using:
    pip install catboost

Contributed by: @AHuzail
"""

import numpy as np
from sklearn.datasets import fetch_california_housing
from sklearn.metrics import mean_squared_error
from sklearn.model_selection import train_test_split
from catboost import CatBoostRegressor


def data_handling() -> tuple:
    """
    Loads and handles the California Housing dataset (replacement for deprecated Boston dataset).

    Returns:
        tuple: A tuple of (features, target), where both are numpy arrays.

    Example:
    >>> features, target = data_handling()
    >>> isinstance(features, np.ndarray)
    True
    >>> isinstance(target, np.ndarray)
    True
    >>> features.shape
    (20640, 8)
    >>> target.shape
    (20640,)
    """
    housing = fetch_california_housing()
    features = housing.data
    target = housing.target
    return features, target


def catboost_regressor(features: np.ndarray, target: np.ndarray) -> CatBoostRegressor:
    """
    Trains a CatBoostRegressor using the provided features and target values.

    Args:
        features (np.ndarray): The input features for the regression model.
        target (np.ndarray): The target values for the regression model.

    Returns:
        CatBoostRegressor: A trained CatBoost regressor model.

    Example:
    >>> features, target = data_handling()
    >>> model = catboost_regressor(features, target)
    >>> isinstance(model, CatBoostRegressor)
    True
    """
    regressor = CatBoostRegressor(iterations=100, learning_rate=0.1, depth=6, verbose=0)
    regressor.fit(features, target)
    return regressor


def main() -> None:
    """
    Main function to run the CatBoost Regressor example.

    It loads the data, splits it into training and testing sets,
    trains the regressor on the training data, and evaluates its performance
    on the test data.

    Example:
    >>> main()
    Mean Squared Error on Test Set:
    """
    # Load and split the dataset
    features, target = data_handling()
    x_train, x_test, y_train, y_test = train_test_split(
        features, target, test_size=0.25, random_state=42
    )

    # Train CatBoost Regressor
    regressor = catboost_regressor(x_train, y_train)

    # Predict on the test set
    predictions = regressor.predict(x_test)

    # Evaluate the performance using Mean Squared Error
    mse = mean_squared_error(y_test, predictions)
    print(f"Mean Squared Error on Test Set: {mse:.4f}")


if __name__ == "__main__":
    import doctest

    doctest.testmod(verbose=True)
    main()