如何从 statsmodel 中总结逻辑回归

【问题标题】：How to Summarize a Logistic Regression from statsmodel如何从 statsmodel 中总结逻辑回归
【发布时间】：2020-05-29 12:09:26
【问题描述】：

我写了下面的代码，但是我想从statsmodel做一个总结，有人可以帮我吗？

谢谢。

from sklearn.model_selection import train_test_split
from sklearn.linear_model import LogisticRegression

X = df[['age_over_65', 'female_perc', 'foreign_born_perc','bachelors_perc', 'household_income']]
y = df['winner']
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.33, random_state=42)

logmodel = LogisticRegression(solver='lbfgs')
logmodel.fit(X_train,y_train)
model = logmodel.predict(X_test)

【问题讨论】：

标签： python data-science logistic-regression statsmodels

【解决方案1】：

Sci-Kit learn 专注于机器学习性能而非统计推断。

如果您想查看 logit 模型的摘要结果，最好使用 statsmodels

下面的示例代码。

import statsmodels.formula.api as smf
X = df[['age_over_65', 'female_perc', 'foreign_born_perc','bachelors_perc', 'household_income']]
y = df['winner']
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.33, random_state=42)
XVARS = ['age_over_65', 'female_perc', 'foreign_born_perc','bachelors_perc', 'household_income']
target = 'winner'
model = smf.logit(formula=f"{target} ~ {' + '.join(XVARS)}", data=df.loc[X_train.index])
logmodel = model.fit()
logmodel.summary2()

#to save in a text file.

with open('logit_results.txt'), 'w') as text_file:
    print(logmodel.summary2(), file=text_file)

【讨论】：