このチャレンジでは、同じ住宅データセットを使用します。ただし、今回は2つの特徴量、すなわち家の築年数と面積（カラム `'age'` および `'square_feet'`）が含まれています。

import pandas as pd

df = pd.read_csv('https://codefinity-content-media.s3.eu-west-1.amazonaws.com/b22d1166-efda-45e8-979e-6c3ecfc566fc/houseprices.csv')
print(df.head())

この課題では、`OLS` クラスを使用して重回帰モデルを構築します。また、各特徴量のp値を確認するためにサマリーテーブルを出力します。

import unittest
import numpy as np
import pandas as pd
import statsmodels.api as sm


# Codefinity helper
def _dynamic_test(test_case, condition, success, fail):
    if condition:
        test_case._testMethodName = success
        test_case.assertTrue(True)
    else:
        test_case._testMethodName = fail
        test_case.fail(fail)


class TestUserCode(unittest.TestCase):

    # 1. Test that X contains the correct columns
    def test_X_contains_correct_columns(self):
        import user_code

        cond = False
        try:
            X = user_code.X
            cond = list(X.columns) == ["age", "square_feet"]
        except Exception:
            cond = False

        _dynamic_test(
            self,
            cond,
            "X correctly contains ['age', 'square_feet'].",
            "Expected X to contain columns: ['age', 'square_feet']."
        )

    # 2. Test preprocessing with sm.add_constant
    def test_X_tilde_add_constant(self):
        import user_code

        cond = False
        try:
            X = user_code.X
            X_tilde = user_code.X_tilde

            cond = (
                isinstance(X_tilde, pd.DataFrame)
                and "const" in X_tilde.columns
                and X_tilde.shape[1] == X.shape[1] + 1
            )
        except Exception:
            cond = False

        _dynamic_test(
            self,
            cond,
            "X_tilde correctly includes a constant column.",
            "Expected X_tilde to be created using sm.add_constant(X)."
        )

    # 3. Model is fitted using OLS
    def test_model_is_fitted(self):
        import user_code
        from statsmodels.regression.linear_model import RegressionResultsWrapper

        cond = isinstance(user_code.model, RegressionResultsWrapper)

        _dynamic_test(
            self,
            cond,
            "model is a fitted OLS RegressionResultsWrapper.",
            "Expected model = sm.OLS(y, X_tilde).fit()"
        )

    # 4. Test X_new_tilde construction
    def test_X_new_tilde(self):
        import user_code
        import numpy as np

        cond = False
        try:
            X_new = user_code.X_new
            X_new_tilde = user_code.X_new_tilde

            expected_shape = (3, 3)  # const + 2 features

            cond = (
                isinstance(X_new_tilde, np.ndarray)
                and X_new_tilde.shape == expected_shape
                and np.allclose(X_new_tilde[:, 1:], X_new)
            )
        except Exception:
            cond = False

        _dynamic_test(
            self,
            cond,
            "X_new_tilde correctly created with sm.add_constant(X_new).",
            "Expected X_new_tilde to include constant column and match X_new values."
        )

    # 5. Test predictions
    def test_predictions(self):
        import user_code
        import numpy as np

        cond = False
        try:
            y_pred = user_code.y_pred
            cond = (
                isinstance(y_pred, np.ndarray)
                and y_pred.shape == (3,)
            )
        except Exception:
            cond = False

        _dynamic_test(
            self,
            cond,
            "y_pred is a NumPy array with correct shape (3,).",
            "Expected y_pred = model.predict(X_new_tilde) to have shape (3,)."
        )

    # 6. Test summary call (using code inspection)
    def test_model_summary_called(self):
        cond = False
        try:
            with open("user_code.py", "r") as f:
                src = f.read()
            cond = "model.summary()" in src.replace(" ", "")
        except Exception:
            cond = False

        _dynamic_test(
            self,
            cond,
            "model.summary() is called.",
            "Expected model.summary() to be printed."
        )


if __name__ == "__main__":
    unittest.main()


test_code.py

すべて正しく実行できていれば、p値はゼロに近い値になります。これは、すべての特徴量がモデルにとって有意であることを意味します。

Pythonを使用した連続的な結果のモデリングおよび予測のための回帰手法を探求します。線形回帰の基礎に基づき、変数間の関係の分析、モデル係数の解釈、実データセットにおける予測性能の評価を行います。


チャレンジ：2つの特徴量を用いた価格予測

解答