Add enhanced model tests for day5 homework

.github/workflows/model_test.yml

@@ -0,0 +1,25 @@
+name: Model Tests
+
+on:
+  push:
+    branches: [ main, day5-homework ]
+  pull_request:
+    branches: [ main ]
+
+jobs:
+  test:
+    runs-on: ubuntu-latest
+    steps:
+    - uses: actions/checkout@v2
+    - name: Set up Python
+      uses: actions/setup-python@v2
+      with:
+        python-version: '3.9'
+    - name: Install dependencies
+      run: |
+        python -m pip install --upgrade pip
+        if [ -f day5/requirements.txt ]; then pip install -r day5/requirements.txt; fi
+    - name: Run tests
+      run: |
+        cd day5/演習3
+        pytest -v tests/

day5/演習3/tests/test_model.py

@@ -171,3 +171,90 @@ def test_model_reproducibility(sample_data, preprocessor):
     assert np.array_equal(
         predictions1, predictions2
     ), "モデルの予測結果に再現性がありません"
+
+def test_model_comparison_with_baseline():
+    """現在のモデルをベースラインモデルと比較"""
+    # ベースラインモデルの読み込み（あらかじめ保存しておく必要あり）
+    baseline_model_path = os.path.join(os.path.dirname(__file__), "../baseline_models/baseline_model.pkl")
+
+    if not os.path.exists(baseline_model_path):
+        pytest.skip("ベースラインモデルが存在しないためスキップします")
+
+    # 現在のモデルの学習
+    data = pd.read_csv(DATA_PATH)
+    X = data.drop("Survived", axis=1)
+    y = data["Survived"].astype(int)
+    X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42)
+
+    preprocessor = ColumnTransformer(
+        transformers=[
+            ("num", Pipeline(steps=[("imputer", SimpleImputer(strategy="median")), ("scaler", StandardScaler())]), ["Age", "Fare"]),
+            ("cat", Pipeline(steps=[("imputer", SimpleImputer(strategy="most_frequent")), ("onehot", OneHotEncoder(handle_unknown="ignore"))]), ["Pclass", "Sex", "Embarked"]),
+        ],
+        remainder="drop",
+    )
+
+    current_model = Pipeline(
+        steps=[
+            ("preprocessor", preprocessor),
+            ("classifier", RandomForestClassifier(n_estimators=100, random_state=42)),
+        ]
+    )
+
+    current_model.fit(X_train, y_train)
+
+    # ベースラインモデルの読み込み
+    with open(baseline_model_path, "rb") as f:
+        baseline_model = pickle.load(f)
+
+    # 両方のモデルで予測
+    baseline_pred = baseline_model.predict(X_test)
+    current_pred = current_model.predict(X_test)
+
+    # 精度の比較
+    baseline_accuracy = accuracy_score(y_test, baseline_pred)
+    current_accuracy = accuracy_score(y_test, current_pred)
+
+    # 現在のモデルはベースラインと同等以上の性能であるべき
+    assert current_accuracy >= baseline_accuracy * 0.95, f"現在のモデル精度({current_accuracy:.4f})がベースライン({baseline_accuracy:.4f})より5%以上低下しています"
+
+def test_detailed_inference_time():
+    """モデルの推論時間詳細テスト（バッチサイズ別）"""
+    # モデルとデータの準備
+    data = pd.read_csv(DATA_PATH)
+    X = data.drop("Survived", axis=1)
+    y = data["Survived"].astype(int)
+    X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42)
+
+    preprocessor = ColumnTransformer(
+        transformers=[
+            ("num", Pipeline(steps=[("imputer", SimpleImputer(strategy="median")), ("scaler", StandardScaler())]), ["Age", "Fare"]),
+            ("cat", Pipeline(steps=[("imputer", SimpleImputer(strategy="most_frequent")), ("onehot", OneHotEncoder(handle_unknown="ignore"))]), ["Pclass", "Sex", "Embarked"]),
+        ],
+        remainder="drop",
+    )
+
+    model = Pipeline(
+        steps=[
+            ("preprocessor", preprocessor),
+            ("classifier", RandomForestClassifier(n_estimators=100, random_state=42)),
+        ]
+    )
+
+    model.fit(X_train, y_train)
+
+    # 異なるバッチサイズでの推論時間テスト
+    batch_sizes = [1, 10, 50, 100]
+    for batch_size in batch_sizes:
+        # バッチサイズに合わせてデータを取得
+        if batch_size <= len(X_test):
+            X_batch = X_test.iloc[:batch_size]
+
+            # 推論時間計測
+            start_time = time.time()
+            model.predict(X_batch)
+            inference_time = time.time() - start_time
+
+            # バッチサイズ1の場合、0.01秒以内、その他は1秒以内であるべき
+            max_time = 0.01 if batch_size == 1 else 1.0
+            assert inference_time < max_time, f"バッチサイズ{batch_size}での推論時間({inference_time:.4f}秒)が長すぎます"