Pipeline 的最后一步应该实现 fit 或者是字符串 'passthrough'答案

【问题标题】：Last step of Pipeline should implement fit or be the string 'passthrough'Pipeline 的最后一步应该实现 fit 或者是字符串 'passthrough'
【发布时间】：2021-10-06 23:13:40
【问题描述】：

我正在尝试构建简单的管道：

from sklearn.linear_model import Lasso
from sklearn.preprocessing import PolynomialFeatures
from sklearn.pipeline import make_pipeline


make_pipeline([
    ('PolynomialFeatures', PolynomialFeatures(include_bias=False)),
    ('Lasso', Lasso(fit_intercept=True, max_iter=1000))])

我得到了错误：

TypeError: Last step of Pipeline should implement fit or be the string 'passthrough'. '[('PolynomialFeatures', PolynomialFeatures(include_bias=False)), ('Lasso', Lasso())]' (type <class 'list'>) doesn't

怎么了？我该如何解决？

【问题讨论】：

标签： machine-learning scikit-learn

【解决方案1】：

您应该使用pipeline 而不是make_pipeline，因为您提供了带有名称的步骤（更多关于此here！）。

from sklearn.linear_model import Lasso
from sklearn.preprocessing import PolynomialFeatures
from sklearn.pipeline import make_pipeline, Pipeline


Pipeline([
    ('PolynomialFeatures', PolynomialFeatures(include_bias=False)),
    ('Lasso', Lasso(fit_intercept=True, max_iter=1000))])

输出：

Pipeline(steps=[('PolynomialFeatures', PolynomialFeatures(include_bias=False)),
                ('Lasso', Lasso())])

【讨论】：