As the title says, I would like to fit a LASSO model on time-series data in R. Instead of k-fold cross-validation, I want to do end of sample cross validation where I can sp