AttributeError: This 'Pipeline' has no attribute 'fit_resample还是这样

--- ### ❗问题再次确认 你仍然遇到以下错误： ``` AttributeError: This 'Pipeline' has no attribute 'fit_resample' ``` 说明：**你使用的 `Pipeline` 实际上是 `sklearn.pipeline.Pipeline`，而不是 `imblearn.pipeline.Pipeline`。** 即使你写了： ```python from imblearn.pipeline import Pipeline as ImbPipeline ``` 但如果在运行时依然报错，说明你的代码中存在**模块导入冲突或覆盖**。 --- ## ✅ 诊断 + 彻底解决方法 我们来一步步排查并彻底解决这个问题。 --- ### 🔍 步骤一：检查是否真的使用了 `imblearn.pipeline.Pipeline` 请你在定义 pipeline 前插入如下调试语句： ```python import imblearn print(imblearn.__version__) ``` 如果你看到类似： ``` AttributeError: module 'imblearn' has no attribute '__version__' ``` 说明你没有正确安装 `imbalanced-learn`！ --- ### 🛠️ 解决方案 A：确保安装了 `imbalanced-learn` 打开终端（命令行），执行： ```bash pip install -U imbalanced-learn ``` 然后重启 Python 内核（Jupyter Notebook 中点击 Kernel > Restart）或重新启动脚本。 --- ### 🧪 解决方案 B：强制验证导入的是 `imblearn` 的 Pipeline 加入如下测试代码： ```python from imblearn.pipeline import Pipeline as ImbPipeline print(ImbPipeline.__module__) # 应该输出：imblearn.pipeline ``` 如果输出是： ``` sklearn.pipeline ``` 那说明你的环境有严重冲突，建议： ```bash pip uninstall imblearn pip install imbalanced-learn ``` ⚠️ 注意：`pip install imblearn` 是旧版本，不推荐使用；应始终使用： ```bash pip install imbalanced-learn ``` --- ### ✅ 解决方案 C：完整无冲突示例代码（复制粘贴即可运行） ```python # 强制检查 imblearn 是否可用 import imblearn from imblearn.pipeline import Pipeline as ImbPipeline print(f"Using imblearn version: {imblearn.__version__}") print(f"ImbPipeline type: {ImbPipeline.__module__}") import pandas as pd from imblearn.over_sampling import BorderlineSMOTE from imblearn.under_sampling import RandomUnderSampler from sklearn.preprocessing import StandardScaler from sklearn.feature_selection import SelectKBest, f_classif # 1. 读取数据 train_df = pd.read_excel(r"C:\Users\86158\Desktop\训练集.xlsx") test_df = pd.read_excel(r"C:\Users\86158\Desktop\测试集.xlsx") X_train = train_df.drop(columns=['target']) y_train = train_df['target'] X_test = test_df.drop(columns=['target']) y_test = test_df['target'] # 2. 特征选择 selector = SelectKBest(score_func=f_classif, k=10) X_train_selected = selector.fit_transform(X_train, y_train) X_test_selected = selector.transform(X_test) # 3. 构建 ImbPipeline pipeline = ImbPipeline([ ('smote', BorderlineSMOTE(random_state=42)), ('under', RandomUnderSampler(random_state=42)), ('scaler', StandardScaler()) ]) # ✅ 执行 fit_resample X_resampled, y_resampled = pipeline.fit_resample(X_train_selected, y_train) print("✅ fit_resample 成功完成！") ``` --- ### ✅ 如果这段代码可以运行，则说明一切正常！ --- ###