Dear all, I'm using package "ridge" to deal with multicollinearity. It's been convenient to automatically choose lambda. However, how do I tell whether the OLS results have been improved after applying ridge regression? I only notice that more variables become statistically significant and some variables' std.errors have been decreased. What is the code to compute R squared for linearRidge() under package "ridge"? Thanks a lot! Hermia [[alternative HTML version deleted]]
Dear all, I'm using package "ridge" to deal with multicollinearity. It's been convenient to automatically choose lambda. However, how do I tell whether the OLS results have been improved after applying ridge regression? I only notice that more variables become statistically significant and some variables' std.errors have been decreased. What is the code to compute R squared for linearRidge() under package "ridge"? Thanks a lot! Hermia [[alternative HTML version deleted]]