Not able to calculate BIC in bidirectional stepwise regression in R

Asked Oct 14 '22 at 18:21

Active Oct 15 '22 at 21:53

Viewed 158 times

I am trying to perform stepwise regression with direction as "both" based on BIC in r.

But everytime is performing based on AIC only

Following is my code

train <- read.csv("Regression_train.csv", header=TRUE, 
    stringsAsFactors=FALSE)
test <- read.csv("Regression_test.csv", header=TRUE, 
    stringsAsFactors=FALSE)
fit.all <- lm(joyjoy ~ ., train)
null <- lm(joyjoy ~ 1, train )
step(null, scope=list(lower=null, upper=fit.all), 
    criterion = "BIC", direction="both", k = log(n))

I don't know why it is performing stepwise bidirectional based on AIC only even though after specifying BIC

Can someone please help here?

edited Oct 15 '22 at 21:53

kjetil b halvorsen

77,844

asked Oct 14 '22 at 18:21

jigar somaiya

Questions about how to get R to do what you want are off topic here. They can be on topic on [SO], but require a reproducible example (among other things). – gung - Reinstate Monica Oct 14 '22 at 19:16
1

FWIW, you almost certainly should not use stepwise selection. It may help you to read my answer to Algorithms for automatic model selection. – gung - Reinstate Monica Oct 14 '22 at 19:18

Not able to calculate BIC in bidirectional stepwise regression in R

0 Answers0