How to specify a specific level as a reference level in GNU R while applying (generalized) linear regression?
A dataframe dt
is given:
y<-rpois(40,1) x1<-gl(4,10) x2<-gl(4,1,40) dt <- data.frame(cbind(y,x1,x2))
Applying poisson regression:
m1<-glm(y~factor(x1)+factor(x2),data=dt) summary(m1)
The output shows x1=1 and x2=1 are both reference level defined by R automatically. To switch the reference level from x1=1 to x1=2, the function relevel
could be used:
m2<-glm(y~relevel(factor(dt$x1),"2")+factor(x2),data=dt) summary(m2)
No comments:
Post a Comment