如何解决Logit二进制测试解释
是否有可能在 Logit 中某个二进制变量的值是 1 但不是 0?我很困惑我应该如何解释它:我应该分别用这些值估计模型吗? 如何使用变量 ex 的常数值估计 logit。吸烟 =1?
现在看起来像这样:
logit death_event anaemia creatinine_phosphokinase diabetes ejection_fraction high_blood_pressure platelets serum_creatinine serum_sodium sex smoking time
Logistic regression Number of obs = 299
LR chi2(12) = 155.79
Prob > chi2 = 0.0000
Log likelihood = -109.77707 Pseudo R2 = 0.4151
------------------------------------------------------------------------------------------
death_event | Odds Ratio Std. Err. z P>|z| [95% Conf. Interval]
-------------------------+----------------------------------------------------------------
age | 1.048561 .0165679 3.00 0.003 1.016586 1.081542
anaemia | .9925574 .3578061 -0.02 0.983 .4896737 2.011891
creatinine_phosphokinase | 1.000222 .000178 1.25 0.212 .9998735 1.000571
diabetes | 1.156213 .4060488 0.41 0.679 .5809054 2.301283
ejection_fraction | .9262024 .0151241 -4.69 0.000 .8970291 .9563245
high_blood_pressure | .9024162 .3237029 -0.29 0.775 .4467608 1.822799
platelets | .9999988 1.89e-06 -0.64 0.525 .9999951 1.000003
serum_creatinine | 1.946618 .3532967 3.67 0.000 1.363936 2.778225
serum_sodium | .9352129 .0371608 -1.69 0.092 .865143 1.010958
sex | .5864558 .2427446 -1.29 0.197 .2605599 1.319967
smoking | .9865984 .4070881 -0.03 0.974 .4394601 2.214937
time | .9791753 .0029516 -6.98 0.000 .9734073 .9849775
_cons | 26500.77 149903.5 1.80 0.072 .4058606 1.73e+09
------------------------------------------------------------------------------------------
. test smoking = 1
( 1) [death_event]smoking = 1
chi2( 1) = 6.03
Prob > chi2 = 0.0140
. test smoking = 0
( 1) [death_event]smoking = 0
chi2( 1) = 0.00
Prob > chi2 = 0.9739
版权声明:本文内容由互联网用户自发贡献,该文观点与技术仅代表作者本人。本站仅提供信息存储空间服务,不拥有所有权,不承担相关法律责任。如发现本站有涉嫌侵权/违法违规的内容, 请发送邮件至 dio@foxmail.com 举报,一经查实,本站将立刻删除。