求助：解决'Pipeline'对象无'fit_resample'属性的AttributeError问题

我参考了一篇关于不平衡分类与管道的文档，尝试在数据集上实现以下代码：

import numpy as np
import pandas as pd
from collections import Counter
from sklearn.preprocessing import LabelEncoder,OneHotEncoder
from imblearn.over_sampling import SMOTE
from imblearn.under_sampling import RandomUnderSampler
from sklearn.pipeline import  Pipeline
from sklearn.naive_bayes import GaussianNB
data =pd.read_csv('aug_train.csv')
data.drop('id',axis=1,inplace=True)
print(data.info())
print(data.select_dtypes(include='object').columns.tolist())
data[data.select_dtypes(include='object').columns.tolist()]=data[data.select_dtypes(include='object').columns.tolist()].apply(LabelEncoder().fit_transform)
print(data.head())
#print(data['Response'].value_counts())
mymodel =GaussianNB()
y =data['Response'].values
print(Counter(y))
X =data.drop('Response',axis=1).values
#X,y =SMOTE().fit_resample(X,y)
#mymodel.fit(X,y)
#print(mymodel.score(X,y))
#print(Counter(y))
over = SMOTE(sampling_strategy=0.1)
under = RandomUnderSampler(sampling_strategy=0.5)
steps = [('o', over), ('u', under)]
pipeline = Pipeline(steps=steps)
# transform the dataset
X, y = pipeline.fit_sample(X, y)

执行上述代码中X, y = pipeline.fit_sample(X, y)行时，出现错误：AttributeError: 'Pipeline' object has no attribute 'fit_resample'，请问该如何修复此问题？

问题根源有两个：

• 导入了错误的Pipeline类：sklearn原生的Pipeline不支持重采样操作的fit_resample方法，必须使用imblearn库提供的专用Pipeline。
• 方法名拼写错误：imblearn的Pipeline对应的正确方法是fit_resample，而非代码中的fit_sample。

具体修复步骤：

1. 替换Pipeline的导入语句：

   # 原导入
   # from sklearn.pipeline import Pipeline
   # 替换为
   from imblearn.pipeline import Pipeline
   

2. 修正方法调用：

   # 原代码
   # X, y = pipeline.fit_sample(X, y)
   # 修正为
   X, y = pipeline.fit_resample(X, y)
   

修复后的完整关键代码片段：

from imblearn.pipeline import Pipeline

over = SMOTE(sampling_strategy=0.1)
under = RandomUnderSampler(sampling_strategy=0.5)
steps = [('o', over), ('u', under)]
pipeline = Pipeline(steps=steps)
# 转换数据集
X, y = pipeline.fit_resample(X, y)

内容的提问来源于stack exchange，提问作者dato datuashvili