sklearn 中 pipeline 或 LabelBinariy出现 'fit_transform() takes 2 positional arguments but 3 were given'

原创

于 2018-09-25 23:49:29 发布 · 4.2k 阅读

2 ·

CC 4.0 BY-SA版权

文章标签：

#Python #sklearn #pipeline

在学习《Hands-On Machine Learning with Scikit-Learn and TensorFlow》时，遇到一个错误：'fit_transform() takes 2 positional arguments but 3 were given'。问题出在sklearn 0.19版本更新了LabelBinarizer，现在只接受两个参数。解决方案是分开处理文本和数值属性，再使用numpy concatenate合并数组。

摘要生成于 C知道，由 DeepSeek-R1 满血版支持，前往体验 >

在学习OReilly.Hands-On.Machine.Learning.with.Scikit-Learn.and.TensorFlow.2017.3时，执行以下代码会出错：


from sklearn.pipeline import FeatureUnion
from sklearn.pipeline import Pipeline
from sklearn.preprocessing import StandardScaler
num_attribs = ['longitude',
 'latitude',
 'housing_median_age',
 'total_rooms',
 'total_bedrooms',
 'population',
 'households',
 'median_income']

cat_attribs = ["ocean_proximity"]
# 在sklearn 0.19下，由于LabelBinarizer的重写而导致代码失效
num_pipeline = Pipeline([
    ('selector', DataFrameSelector(num_attribs)),
    ('imputer', Imputer(strategy="median")