sklearn：使用 ColumnTransformer 链接多个变压器

Question

如何使用 ColumnTransformer API将多个转换器应用到单个pandas DataFrame 列？

例如，我想取立方根，然后标准化 DataFrame 列中的值：

df = pd.DataFrame(

  np.array([[1., 2., 3.], [4., 5., 6.], [7., 8., 9.]]),

  columns=['a', 'b', 'c']

)

transformer = ColumnTransformer(

  [

    ('root3_std', StandardScaler() + FunctionTransformer(np.cbrt), 'a') <-- pseudocode

  ],

  remainder='passthrough'

)

如果我写

transformer = ColumnTransformer(

  [

    ('root3', FunctionTransformer(np.cbrt), 'a'),

    ('standardize', StandardScaler(), 'a')

  ],

  remainder='passthrough'

)

我得到两列独立的列，一列包含立方根，另一列包含标准化原始值。如何一次性应用两个变压器？

Answer 1

from sklearn.pipeline import Pipelineimport pandas as pdimport numpy as npfrom sklearn.preprocessing import FunctionTransformer, StandardScalerdf = pd.DataFrame(np.array([[1., 2., 3.], [4., 5., 6.], [7., 8., 9.]]),columns=['a', 'b', 'c'])pipe = Pipeline([('function_transformer', FunctionTransformer(np.cbrt)),&nbsp;&nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp;('standard_scalar', StandardScaler())])pipe.fit_transform(df[['a']])#oparray([[-1.32381804],&nbsp; &nbsp;[ 0.23106179],&nbsp; &nbsp;[ 1.09275626]])