ridgregextra focuses on finding the ridge parameter value k which makes the VIF values closest to 1 while keeping them above 1 as stressed “Applied Linear Statistical Models” (Kutner et al., 2004). The package includes the ridgereg_k function, presents a system that automatically determines the k value in a certain range defined by the user and provides detailed ridge regression results. ridgereg_k also provides ridge regression tables (VIF, MSE, R2, Beta, Stdbeta) using vif_k function for k ridge parameter values generated between certain lower and upper bound values.
In addition, the ridge_reg function provides users the ridge regression results for a manually entered k value. Finally ridgregextra provides three sets of graphs consisting k versus VIF values, regression coefficents and standard errors of them.
ridgregextra was presented for the first time in “Why R? Turkey 2022” conference.
ridgregextra from CRANinstall.packages("ridgregextra")ridgregextra development versionPlease make sure that you installed devtools package.
If you would like to install dev version of the package, please use following command.
devtools::install_github(filizkrdg/ridgregextra)You can use isdals package to have example data to test ridgregextra package. isdals package is being installed, while you are installing ridgregextra package, so you don’t have to install the package again.
library(isdals)
data(bodyfat)
x=bodyfat[,-1]
y=bodyfat[,1]ridgereg_k function to get coefficients by using alternative approach to traditional ridge regression techniques.ridgereg_k(x,y,0,1)
You can use mctest package to have example data to test ridgregextra package. mctest package is being installed, while you are installing ridgregextra package, so you don’t have to install the package again.
library("mctest")
x=Hald[,-1]
y=Hald[,1]ridgereg_k function to get coefficients by using alternative approach to traditional ridge regression techniques.ridgereg_k(x,y,0,1)For any questions please contact: