次のデータセット（**2つの特徴量**を持つ）を使用:

import pandas as pd

df = pd.read_csv('https://codefinity-content-media.s3.eu-west-1.amazonaws.com/b71ff7ac-3932-41d2-a4d8-060e24b00129/circles.csv')
print(df.head())

以下のコードを実行し、得られた**散布図**を確認すると、このデータセットが**線形分離不可能**であることが分かる:

import pandas as pd
import matplotlib.pyplot as plt

df = pd.read_csv('https://codefinity-content-media.s3.eu-west-1.amazonaws.com/b71ff7ac-3932-41d2-a4d8-060e24b00129/circles.csv')
plt.scatter(df['X1'], df['X2'], c=df['y'])
plt.show()

このデータに対して**交差検証**を用いて、単純なロジスティック回帰を評価。

import pandas as pd
import matplotlib.pyplot as plt
from sklearn.preprocessing import StandardScaler
from sklearn.linear_model import LogisticRegression
from sklearn.model_selection import cross_val_score

df = pd.read_csv('https://codefinity-content-media.s3.eu-west-1.amazonaws.com/b71ff7ac-3932-41d2-a4d8-060e24b00129/circles.csv')
X = df[['X1', 'X2']]
y = df['y']

X = StandardScaler().fit_transform(X)
lr = LogisticRegression().fit(X, y)

y_pred = lr.predict(X)
plt.scatter(df['X1'], df['X2'], c=y_pred)
plt.show()

print(f'Cross-validation accuracy: {cross_val_score(lr, X, y).mean():.2f}')

ご覧のとおり、通常のロジスティック回帰はこのタスクには適していません。多項式回帰を使用することで、モデルの性能向上が期待できます。さらに、`GridSearchCV` を利用することで、より高い精度のための最適な `C` パラメータを見つけることができます。

このタスクでは `Pipeline` クラスも使用します。これは**前処理ステップの連続**と考えることができます。その `.fit_transform()` メソッドは、パイプライン内の各ステップに対して順番に `.fit_transform()` を適用します。

import typing
import unittest
import importlib


def _dynamic_test(test_case, condition, success_message, failure_message):
    if condition:
        test_case._testMethodName = success_message
        test_case.assertTrue(True, success_message)
    else:
        test_case._testMethodName = failure_message
        test_case.fail(failure_message)


class TestUserCode(unittest.TestCase):

    def test_pipe_is_declared(self):
        import user_code
        _dynamic_test(
            self,
            hasattr(user_code, 'pipe'),
            "The `pipe` variable is declared.",
            "Expected the `pipe` variable to be declared."
        )

    def test_pipe_contains_polynomial_and_scaler(self):
        import user_code
        from sklearn.preprocessing import StandardScaler, PolynomialFeatures
        from sklearn.pipeline import Pipeline

        expected_value = [PolynomialFeatures(include_bias=False), StandardScaler()]

        variable = 'pipe'
        actual_value = getattr(user_code, variable, None)
        condition = False
        if actual_value is None:
            failure_message = f"The `{variable}` variable is not declared."
        elif isinstance(actual_value, Pipeline):
            if isinstance(actual_value.steps, list):
                if len(actual_value.steps) != len(expected_value):
                    failure_message = f'Expected the number of steps of the `pipe` to be `{len(expected_value)}`, but got `{len(actual_value.steps)}`.'
                else:
                    try:
                        steps = [obj for _, obj in actual_value.steps]
                        condition = all(
                            steps[i].get_params() == expected_value[i].get_params() for i in range(len(steps)))
                    except TypeError:
                        steps = actual_value.steps
                        condition = False

                    failure_message = f"Expected the steps of the `{variable}` to be `{expected_value}`, but got `{steps}`."
            else:
                failure_message = "The argument of the `Pipeline` is not a `list`."
        else:
            failure_message = f"`{variable}` is not a `Pipeline`."

        _dynamic_test(
            self,
            condition,
            f"`{variable}` contains correct steps.",
            failure_message
        )

    def test_param_grid_is_declared(self):
        import user_code

        _dynamic_test(
            self,
            hasattr(user_code, 'param_grid'),
            "The `param_grid` variable is declared.",
            "Expected `param_grid` to be declared."
        )

    def test_param_grid_is_correct(self):
        import user_code
        expected_value = {'C': [0.01, 0.1, 1, 10, 100]}

        variable = 'param_grid'
        actual_value = getattr(user_code, variable, None)
        if actual_value is None:
            condition = False
            failure_message = f"The `{variable}` variable is not declared."
        elif isinstance(actual_value, dict):
            condition = actual_value == expected_value
            failure_message = f"Expected `{variable}` to contain `{expected_value}`, but got `{actual_value}`."
        else:
            condition = False
            failure_message = f"`{variable}` is not a `dict`."

        _dynamic_test(
            self,
            condition,
            f"`{variable}` contains the correct values.",
            failure_message
        )

    def test_grid_cv_is_declared(self):
        import user_code

        variable = 'grid_cv'
        _dynamic_test(
            self,
            hasattr(user_code, variable),
            f"The `{variable}` variable is declared.",
            f"Expected `{variable}` to be declared."
        )

    def test_grid_cv_is_correct(self):
        import user_code
        from sklearn.model_selection import GridSearchCV
        from sklearn.linear_model import LogisticRegression

        param_grid = {'C': [0.01, 0.1, 1, 10, 100]}

        variable = 'grid_cv'
        actual_value = getattr(user_code, variable, None)
        if actual_value is None:
            condition = False
            failure_message = f"The `{variable}` variable is not declared."
        elif isinstance(actual_value, GridSearchCV):
            condition = isinstance(actual_value.estimator, LogisticRegression) and actual_value.param_grid == param_grid and actual_value.cv is None
            failure_message = f"Expected `{variable}` to be a `GridSearchCV` with `estimator=LogisticRegression()`, `param_grid={param_grid}`, and `cv=None`, but got `estimator={actual_value.estimator}`, `param_grid={actual_value.param_grid}`, and `cv={actual_value.cv}`."
        else:
            condition = False
            failure_message = f"`{variable}` is not a `GridSearchCV`."

        _dynamic_test(
            self,
            condition,
            "`grid_cv` is a `GridSearchCV` with `estimator=lr`, `param_grid=param_grid`, and `cv=None`.",
            failure_message
        )

    def test_best_score_is_correct(self):
        import user_code
        expected_value = 0.9120000000000001

        variable = 'best_score'
        actual_value = getattr(user_code, variable, None)
        if actual_value is None:
            condition = False
            failure_message = f"The `{variable}` variable is not declared."
        elif isinstance(actual_value, float):
            condition = actual_value == expected_value
            failure_message = f"Expected `{variable}` to contain `{expected_value}`, but got `{actual_value}`."
        else:
            condition = False
            failure_message = f"`{variable}` is not a `float`."

        _dynamic_test(
            self,
            condition,
            f"`{variable}` is correct.",
            failure_message
        )


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

test_main.py

Pythonを使用したカテゴリカルな結果を予測するための分類手法を紹介します。分類モデルの構築、トレーニング、評価、決定境界の解釈、一般的なアルゴリズムの実世界データセットへの適用に焦点を当てます。


チャレンジ：分離不可能なデータの分類

解答