October meeting: Setting up your own Shiny server & Managing courses with GitHub

This is a past event

14 people went

Location image of event venue


As usual, the meeting will be open for all to attend, and newcomers / beginners are very welcome! After the talks we’ll be testing out a new pub (details below), so do come along!

Our first speaker is Colin Rundel, Lecturer in Mathematical Sciences within the School of Mathematics (UoE), and Assistant Professor of the Practice in Statistical Science at Duke University. Colin will be focusing on how to use GitHub for teaching:

# ghclass: an R package for managing classes with GitHub

In this talk I will present the details of the ghclass package which has been developed as a tool for managing statistical and data science courses with a computation component. The package is designed to automate the process of distributing and collecting assignments via git repositories hosted on GitHub. As part of this talk I will discuss how these tools are being used to teach a large MSc Statistical Computing course here at the University of Edinburgh. I will provide a live demo of the package’s core functionality as well as demonstrate some of the newer features which focus on automated feedback and assessment.

Our second speaker is Greg Sutcliffe. Greg is a Community Data Scientist at Red Hat, focusing mainly on supporting the Ansible and Foreman communities. He’s interested how data can be used to guide their governance, particularly at scale. He will be discussing:

# Running your own Shiny Server

Is the free tier of Shinyapps.io too restrictive? Or you need something internal perhaps? If, like me, you’ve got CPU time and no budget, running your own Shiny server is a compelling option. In this demo, we’ll take a clean Debian virtual machine, get Shiny Server up and running, add an Apache frontend, and throw in some LetsEncrypt for SSL goodness. We might even have time to look into how to deploy apps to it from within RStudio.

See you there!