如何得到一个回归方程的异方差-稳健标准误?

heteroskedasticity robust standard error.
有没有这样一个函数in R/?

R一下 “heteroskedasticity robust standard error” 就知道...
http://www.rseek.org

1. D<-read.table('meap00_01.raw')
   
2. names(D)<-c('dcode','bcode','math4','read4','lunch','enroll','exppp','lenroll','lexppp')
   
3. attach(D)
   
4. 
   
5. lm1<-lm(math4~lunch+lenroll+lexppp,data=D)
   
6. summary(lm1)
   
7. 
   
8. lm11<-lm(lunch~lenroll+lexppp)
   
9. se1<-sqrt(t(lm11$residual^2)%*%lm1$residual^2/(sum(lm11$residual^2))^2)
   
10. se1
    
11. lm12<-lm(lenroll~lunch+lexppp)
    
12. se2<-sqrt(t(lm12$residual^2)%*%lm1$residual^2/(sum(lm12$residual^2))^2)
    
13. se2
    
14. lm13<-lm(lexppp~lenroll+lunch)
    
15. se3<-sqrt(t(lm13$residual^2)%*%lm1$residual^2/(sum(lm13$residual^2))^2)
    
16. se3
    
17. 
    
18. coeftest(lm1, vcov=sandwich)

复制代码

coeftest(lm1, vcov=sandwich)
自己google到了

太感谢了！