使用不同的数据名称在R中运行Wald测试(逻辑回归)
各位,我要用数据进行逻辑回归。 这是我的密码:使用不同的数据名称在R中运行Wald测试(逻辑回归),r,R,各位,我要用数据进行逻辑回归。 这是我的密码: library(lmtest) data <- read.csv("C:/Users/user/Desktop/SAFETY.csv",header=TRUE)(your directory) data$Type<-factor(data$Type) data$Size<-factor(data$Size) data$Region<-factor(data$Region) mylogit<- glm(Unsafe~S
library(lmtest)
data <- read.csv("C:/Users/user/Desktop/SAFETY.csv",header=TRUE)(your directory)
data$Type<-factor(data$Type)
data$Size<-factor(data$Size)
data$Region<-factor(data$Region)
mylogit<- glm(Unsafe~Size+Weight+Region, data=data,family = "binomial")
waldtest(mylogit,test="Chisq")
库(lmtest)
数据我无法在R3.2.2和lmtest
0.9.34上重现此问题(请参见末尾的sessionInfo()
输出)
使用此代码:
library(lmtest)
dfS = read.csv("Data/SAFETY.csv")
logitS = glm(Unsafe ~ factor(Size) + Weight +
Region, data = dfS, family = binomial())
summary(logitS)
waldtest(logitS, test = "Chisq")
我能够制作:
> summary(logitS)
Call:
glm(formula = Unsafe ~ factor(Size) + Weight + Region, family = binomial(),
data = dfS)
Deviance Residuals:
Min 1Q Median 3Q Max
-1.7970 -0.5866 -0.3119 0.7864 2.2018
Coefficients:
Estimate Std. Error z value Pr(>|z|)
(Intercept) 2.7285 1.3949 1.956 0.05046 .
factor(Size)2 -2.0200 0.6246 -3.234 0.00122 **
factor(Size)3 -2.6785 0.8810 -3.040 0.00236 **
Weight -0.6678 0.4589 -1.455 0.14559
RegionN America -0.3775 0.5624 -0.671 0.50203
---
Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
(Dispersion parameter for binomial family taken to be 1)
Null deviance: 119.249 on 95 degrees of freedom
Residual deviance: 84.004 on 91 degrees of freedom
AIC: 94.004
Number of Fisher Scoring iterations: 5
> waldtest(logitS, test = "Chisq")
Wald test
Model 1: Unsafe ~ factor(Size) + Weight + Region
Model 2: Unsafe ~ 1
Res.Df Df Chisq Pr(>Chisq)
1 91
2 95 -4 23.987 8.035e-05 ***
---
Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
我的建议是重新启动R会话并重新加载所有必需的包。如果需要,请执行rm(list=ls())
(注意:这将从工作区中删除所有对象)
我无法在R3.2.2和lmtest
0.9.34上重现此问题(请参见末尾的sessionInfo()
输出)
使用此代码:
library(lmtest)
dfS = read.csv("Data/SAFETY.csv")
logitS = glm(Unsafe ~ factor(Size) + Weight +
Region, data = dfS, family = binomial())
summary(logitS)
waldtest(logitS, test = "Chisq")
我能够制作:
> summary(logitS)
Call:
glm(formula = Unsafe ~ factor(Size) + Weight + Region, family = binomial(),
data = dfS)
Deviance Residuals:
Min 1Q Median 3Q Max
-1.7970 -0.5866 -0.3119 0.7864 2.2018
Coefficients:
Estimate Std. Error z value Pr(>|z|)
(Intercept) 2.7285 1.3949 1.956 0.05046 .
factor(Size)2 -2.0200 0.6246 -3.234 0.00122 **
factor(Size)3 -2.6785 0.8810 -3.040 0.00236 **
Weight -0.6678 0.4589 -1.455 0.14559
RegionN America -0.3775 0.5624 -0.671 0.50203
---
Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
(Dispersion parameter for binomial family taken to be 1)
Null deviance: 119.249 on 95 degrees of freedom
Residual deviance: 84.004 on 91 degrees of freedom
AIC: 94.004
Number of Fisher Scoring iterations: 5
> waldtest(logitS, test = "Chisq")
Wald test
Model 1: Unsafe ~ factor(Size) + Weight + Region
Model 2: Unsafe ~ 1
Res.Df Df Chisq Pr(>Chisq)
1 91
2 95 -4 23.987 8.035e-05 ***
---
Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
我的建议是重新启动R会话并重新加载所有必需的包。如果需要,请执行rm(list=ls())
(注意:这将从工作区中删除所有对象)
在这个包中,df
代表自由度。这就是它不起作用的原因。如果您使用df1
或其他任何名称重命名数据,它将起作用。在该包中,df
表示自由度。这就是它不起作用的原因。如果您使用df1
或其他方法重命名数据,则该功能将起作用。