Dear R-Experts,
Here is my reproducible R code to get the Mean squared error of GAM and MARS
after I = 50 iterations/replications.
If I want to get the 95% bootstrap CIs around the MSE of GAM and around the MSE
of MARS, how can I complete/modify my R code ?
Many thanks for your precious help.
##################
library(mgcv)
library(earth)
my.experiment <- function() {
n<-500
x <-runif(n, 0, 5)
z <- rnorm(n, 2, 3)
a <- runif(n, 0, 5)
y_model <- 0.1*x^3 - 0.5*z^2 - a + x*z + x*a + 3*x*a*z + 10
y_obs <- y_model +c( rnorm(n*0.97, 0, 0.1), rnorm(n*0.03, 0, 0.5) )
gam_model<- gam(y_obs~s(x)+s(z)+s(a))
mars_model<-earth(y_obs~x+z+a)
MSE_GAM<-mean((gam_model$fitted.values - y_model)^2)
MSE_MARS<-mean((mars_model$fitted.values - y_model)^2)
return( c(MSE_GAM, MSE_MARS) )
}
my.data = t(replicate( 50, my.experiment() ))
colnames(my.data) <- c("MSE_GAM", "MSE_MARS")
summary(my.data)
##################
______________________________________________
R-help@r-project.org mailing list -- To UNSUBSCRIBE and more, see
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.
