This session will be a mixture of lectures and short practical sessions using R.
Please make sure you are using the latest version of R (current version is 3.4.1) - the final digit is the minor version number. The essential numbers are the first two. You can check the version of R you are running via
R.version.string
Please install the latest version of RStudio (https://www.rstudio.com/products/rstudio/download/) (or another suitable IDE).
We'll also need a couple of packages
install.packages(c("mvtnorm", "glmnet"))
We'll also use the following function in the lectures
get_data = function(n = 100, total_p = 50, real_p = 15,
train_prop = 0.66,
sd = 1,
seed = NULL) {
if(!is.null(seed)) set.seed(seed)
x = matrix(rnorm(n*total_p), nrow=n, ncol=total_p)
y = apply(x[,1:real_p], 1, sum) + rnorm(n, sd = sd)
train_rows = sample(1:n, train_prop * n)
l = list()
l[["x_train"]] = x[train_rows, ]
l[["x_test"]] = x[-train_rows, ]
l[["y_train"]] = y[train_rows]
l[["y_test"]] = y[-train_rows]
l
}
- Links to be added