Software and Troubleshooting

Goals

To install and setup key software for use in the course. Instructions are below.

IF YOU NEED EXTRA HELP: TAs are available from 9am-10:30am in Mathematics Annex 1100 on Tuesday September 5th.

Note on PC vs Mac

You may encounter some material that implicitly assumes that you are using a Mac. For example, they might point you to a Terminal app, or ask you to use Unix commands like which or ls. Here is a cheatsheet that helps you convert between Windows command line (“DOS”) and Mac command line (“UNIX”). One thing that’s missing is the equivalent to Unix’s which; the closest equivalent in DOS is where.

Step 1: R and RStudio

The main programming language used in the class is R and our main IDE (a text editor with helpful features for writing and running code) is RStudio.

  • Follow Chapter 5 of Jenny Bryan’s “happy git” book to install R and RStudio if you haven’t already.

  • If you have R and/or RStudio but haven’t updated them in a long time – e.g. a year or more – then we recommend that you take the time to update them now. This will make it easier for us to help you troubleshoot throughout the class.

Step 2: Jupyter and R Kernel

You will need another IDE for R (Jupyter Notebook) and an R kernel called ‘IR Kernel’ (needed to run R code within Jupyter) to work on the autograded worksheets. This is because the autograder we use plays well with Jupyter.

The following instructions are adapted from Rich Pauloo.

For PC

  1. Install Anaconda from here. This gives you Jupyter and the Anaconda Prompt command line tool.

  2. Install and setup the IR Kernel.

    • Find the location of R.exe on your computer. (The location probably looks something like C:\Program Files\R\R-4.3.1\bin.)
    • Open the Anaconda Prompt application. Enter the following command into Anaconda Prompt: cd file_path_here, with file_path_here replaced with the location of R.exe that you previously found.
    • Run R from within Anaconda Prompt by entering in R.exe. This opens an R session inside Anaconda Prompt. From here, enter the following commands in sequence: install.packages("IRkernel") and IRkernel::installspec(). This respectively installs IR Kernel and tells Jupyter where to find IR Kernel.

  3. To verify that everything is working as intended: Open Anaconda Prompt and type in jupyter lab. Jupyter Lab should launch and display both a Python and R kernel.

For OSX:

  1. Install Anaconda from here. This gives you Jupyter.

  2. Install and setup the IR Kernel.

  • Open up the Terminal application. Run R from within Terminal by entering in R. This opens an R session inside Terminal. From here, enter the following commands in sequence: install.packages("IRkernel") and IRkernel::installspec(). This respectively installs IR Kernel and tells Jupyter where to find IR Kernel.
  1. To verify that everything is working as intended: Open Jupyter Lab (e.g. by opening up the Anacaonda Navigator application and clicking Launch inside the JupyterLab box). Jupyter Lab should launch and display both a Python and R kernel.

Step 3: Git and Github

Git is a version control software. Github is a cloud-based platform (built on Git) for sharing code.

  • Register a Github account at github.com. See Chapter 4

  • Install git. Chapter 6.

  • Configure git. See Chapter 7: Introduce yourself to git.

  • Get setup with a GitHub Personal Access Token (PAT): this will allow you to sync GitHub with files on your computer.

    • Step 1: Create a PAT

      • Option 1: Go to to https://github.com/settings/tokens and click “Generate token”. Describe the token’s purpose in the Note field, e.g. “personal-macbook” or “vm-for-project-xyz”. Select “repo”, “user”, and “workflow” for scopes.

      • Option 2: From R, execute usethis::create_github_token() (you may have to install the usethis package first by first running install.packages("usethis")). Describe the token’s purpose in the Note field, e.g. “personal-macbook” or “vm-for-project-xyz”.

    • Step 2: From R, execute `gitcreds::gitcreds_set()` and enter the PAT that you just made when prompted. (You may have to install the gitcreds package first by first running install.packages("gitcreds")). That should be it!

    • Step 3 (optional but recommended) Also save the PAT that you just made somewhere secure, like a password manager.

Attribution

By Vincenzo Coia and Jenny Bryan, with inputs from STAT 545 teaching teams.