0

I'm not a frequent poster so I apologize if this format is not correct. If you tell me how to show the data I will make that change. In the meantime, here is the code. The vif() generates very different scores at the k=0 level in lmridge. Why are they different?

library(car)
library(lmridge)

data <- Duncan
ds   <- as.data.frame(scale(data[,2:4]))

m <- lm(prestige ~ income + education, data = ds)
vif(m)


m2 <- lmridge(prestige ~ income + education, data = ds,
              scaling = "scaled",
              K = seq(0,1,0.10))
vif.lmridge(m2)
Regis Maria O'Connor
  • 43
  • 5
  • 1
    in lmridge you are scaling the dataset ds again. You should use the Duncan dataset. Note do not call a variable "data". This is already a function in R that you are overwriting here. – phiver Dec 31 '20 at 11:12
  • Thanks phiver for the advice and the answer. That worked! – Regis Maria O'Connor Jan 03 '21 at 19:09

0 Answers0