### Forward Stepwise Regression in R
out0 = lm(y ~ 1) ### fit the initial model (intercept only)
out  = step(out0,direction="forward",trace=TRUE,scope= y ~ x1 + x2 + x3 + x4 + x5)
summary(out)



### lasso
library(lars)
out = lars(x,y) ### x is a matrix of covariates
summary(out)
plot(out)