sklearn_pandas DataFrameMapper弄乱了OneHotEncoder列名

Question

import pandas as pd
import sklearn.preprocessing
import sklearn_pandas

df = pd.DataFrame({'z': ['a', 'b', 'b', 'c']})

sklearn_pandas.DataFrameMapper([
    (['z'], sklearn.preprocessing.OneHotEncoder(drop='first'))
], df_out=True).fit_transform(df)

输出

   z_x0_b  z_x0_c
0     0.0     0.0
1     1.0     0.0
2     1.0     0.0
3     0.0     1.0

为什么列名称混乱？我不要x0在里面。

相关

• using OneHotEncoder with sklearn_pandas DataFrameMapper
• How to output Pandas object from sklearn pipeline