Advanced Statistics

Please see the course web site for more information on the course.

After completing the setup below, you can get the course files.

Software required

To participate in the course, you will need to bring a laptop with the following software packages. Install them in the order listed below!

• R (version 4.2 or later)
• Developer tools for R (see below)
• Rstudio (recommended)
• Stan (installed via the rstan package, see below)

R developer tools

We will use Stan for fitting some models, which requires a C++ compiler and other tools that can be accessed from within R. Follow the instructions below depending on what operating system you are using.

Mac

1. The easiest way to get developer tools is to install XCode (requires an Apple ID and installation through the Mac App Store). Warning: This is quite a large installation, requiring > 20 GB of disk space, and might take hours to complete.
2. Once XCode is complete, launch the Terminal App (located in /Applications/Utilities/). Run the command sudo xcode-select --install
   This installation will also probably take some time to finsih.
3. Accept the XCode license: In the terminal, type sudo xcodebuild -license accept
4. Install gfortran: information here. You only need to follow the instructions under GNU Fortran Compiler.

Windows

You will need to install rtools for your version of R (either < 4.2 or >= 4.2)

Linux

Be sure you have installed both the r-base and r-base-dev packages. Then follow these instructions to configure your system for Stan.

Installing rstan

Windows users running R 4.2

There have been some issues with Windows, rtools and R 4.2. If you encounter problems, install stan with the following command:

install.packages("rstan", repos = c("https://mc-stan.org/r-packages/", getOption("repos")))

Then proceed with testing the installation as written below.

Everyone else

In most cases, after you install the developer tools above, installing Stan should be relatively simple. Simply run the command install.packages('rstan') from within R, then test the installation following these instructions. If it works then you are good to go! If not, read through the Stan installation guide for troubleshooting steps.