Setup and Software

Overview

The STAT 447 course aims to provide rigorous training in modern programming and data science methods. As in the past, we rely on best-of-breed tools. The two most important ones are RStudio and GitHub which we describe below in more detail.

The first of these two tools is RStudio as the primary coding environment. The second is GitHub as a place for git repositories, “social coding” i.e. communication via issue tickets, pull request, markdown documentation and more.

R and RStudio

There are numerous ways to run RStudio. RStudio can run as a desktop application on your computer, as a server process you access via your browser from a machine you have access to, or in Posit Cloud (formerly RStudio Cloud). All three are availble in some form, and you can, and should, make some choices.

RStudio Desktop

You can most likely install RStudio as a desktop application on your (desktop or laptop) computer. Windows, macOS and Linux are supported. You will also have to install R alongside it, the packages we use and possibly other software. Some students will be at ease with this, others may find it daunting.

It might still be best to try this youself as it is a good learning experience, and to get help from fellow students (or campus resources) when you need it. Course staff cannot be expected to provide individual help on installations. Consider RStudio Cloud or RStudio Server (see next section) as alternatives and fallbacks if you are unsure – but considering installing on your local machine.

RStudio Server

As of Summer 2022, we are offering RStudio for registered students via the machine morrow.stat.illinois.edu. Only students registered in STAT 447 will be able to login. To connect, use your NetId as the username and your usual netid password. It is our intent to provide this as a working solution for all students in the class.

While generously sized in terms of memory and cpus, it is a single server so may at times of peak demand shows its limits. So also consider a local desktop installation (as per the previous paragraph).

This is only our second use of the server so let us if anything is not yet to your expectations and we will try to address it. Course staff can also take care of installing R packages from CRAN for you. You can also assume shell tools, markdown, sqlite3 and duckdb etc to present.

Note that we will not generally support arbitrary other tools and cannot support GPU computing on that machine.

Posit / RStudio Cloud

Previous runs of this course relied on a campus-wide pilot of RStudio Cloud (now called [posit cloud][posit cloud]). As of Fall 2022, this option is no longer available. You can still use Posit Clout (formerlly known as RStudio Cloud) individually by signing up at the [posit.cloud][posit cloud] website. Up to 25 hours/months are available on the ‘Cloud Free’ and for $5 you can get 75 hours in the ‘Cloud Plus’ plan, see the pricing and features info for details. This may be enough as a fallback and potentially helpful in case you cannot install some components on your computer, and cannot use the server we provide.

There are other alternatives: Google Colab, Fly.io, and more. You are free to explore these, of course, but none are formally supported here.

GitHub

In the begining, the course has relied on students using personal logins at github.com without the Single Sign-On (SSO) connections, making coordination with netids more difficult and cumbersome. An interim solution was using GitHub in an on-campus instance with netids, this was also cumbersome (for some) as it disconnected from your usual GitHub setup. But now and as in previous years, we are using GitHub as a campus-wide service via Single-Sign On (SSO) into a course-specific ‘organization’ at GitHub. Because you login using SSO, it maps to your campus identification. The GitHUb account can be an existing one or a new one, it separate from your NetID but by requesting access via SSO we are connecting both.

Action Items: We use a GitHub ‘organization’ (a group of repositories and users) for the course called ‘illinois-stat447’. To join, you need a GitHub account (at the external github.com, not an older campus-hosted one). You then need to sign in at this link using your netid and standard SSO two-factor authentication to join this ‘illinois-stat447’ organization. (If you do not have a GitHub account one screen will be offered to create one.) Once you are in the organization, you can request to join the team ‘students-sp24’ via this link. This gives you access to the course discussion repositories, and more.

If you do have not yet have a GitHub account, you can create one here. Additional information, should you need it, is available here

We will be using three repositories for discussion in issue tickets:

As per the action items above, you must be a member of the ‘students-sp24’ team to access these so follow the link to join the organization, and then follow the link to joint the team ‘students-sp24’.

Box

We will continue to use uofi.app.box.com for slides, course videos, and support files. We will generally provide direct links from this website. Access will require SSO via your netid.

You may need to create an account for this ‘box.com’ application if you have never used it before. No additional setup should be needed.

The content of the video folder is also synchronized with this ClassTranscribe playlist.

Mediaspace

A channel may be created at mediaspace.illinois.edu to provide an alternate access method. Access should also require SSO via your netid.

ClassTranscribe

A playlist at ClassTranscribe offers alternate access to our course videos. Access should also require SSO via your netid.

Canvas

While Compass2g was used in the past, we use Canvas starting with the Fall 2022 term. For this term the course URL is canvas.illinois.edu/courses/30522 but if you are enrolled and log into Canvas it will likely be selected for you.

While Canvas may be new for the course, you may have used Canvas in the past for other courses so it should not be an issue. As with Compass2g we plan to rely pn Canvas mainly for (approximately weekly, and as needed) emails to everybody, for grades, and as general roster. We will likely not use any discussion forums etc as we prefer GitHub Issues for this (see next paragraph).

Discussion Forum

We will once again use GitHub issues in a dedicated repository. This a common practice in real-world projects. In the past, course staff has generally reacted within a few hours of each post, and we will again strive to provide timely follow-up.

We also strongly encourage direct student engagements. If you know how to answer a question of a fellow student, post the answer rather than just wait for the course staff to chime in.