• Title: April 2026: Fake data and R projects
• Speakers: Mike Spencer and Claire Hayman
• Date: Thursday 30 April 2026, 6.00PM - 7.00PM
• Location: Teaching Room 02, Old College The University of Edinburgh
• Register: meetup.com/edinbr/events/314336153

Mike Spencer: Making fake data for testing

Ever needed to use real data where you shouldn’t? Maybe you had external collaborators who weren’t cleared to work with live data, or wanted to do a public talk but couldn’t share private information. Come to this session to learn about making good enough data for your purpose. Turbo charge your development with safe data for prototypes and more.

Mike is an independent data consultant who supports organisations with research, engineering and hands-on data maturity projects. He’s a Fellow of the Royal Statistical Society, Member of the Society of Research Software Engineers and also one of the EdinbR conveners. Contact Mike through linkedin or mastodon.

Slides: https://codeberg.org/mikerspencer/workshops/src/branch/main/fake_data/edinbr_slides.pdf

Claire Hayman & Mike Spencer: setting up R for your project

In our second session we’re trying something different - crowd sourcing a great setup for R projects. This will be perfect for those new to R and also old hands. Be prepared to share your wisdom and have preconceptions challenged!

Claire Hayman, Surveillance Information & Data Coordinator at the Animal and Plant Health Agency.

Outcome

We had a really positive and engaging discussion with contribution from all.

• Organise panes in RStudio (e.g. console on the right)
• Enable the native pipe
• Enable rainbow parentheses
• Highlight function calls
• Disable workspace and history retention
• Set font to fierer, which makes pipe prettier
• Git setup on the system, follow happy git with R
• Setup a Git repository on the remote first, then clone (example remotes are Codeberg, Github, Bitbucket, Gitlab, Forgejo, etc.)
• Start an RStudio project
• Consider controlling the environment. However, renv only controls packages, but not wider software (R, system dependencies). Using containers is a more robust approach (advanced)
• Use .gitignore to keep data out of version control
• Fetch remote data with a script that checks if a file exists
• Git vaccinate (from usethis) to help shield from accidental leaks.

Slides from Claire and Mike: https://github.com/EdinbR/edinbr-talks/blob/master/2026-04-30/2026-04_project-setup-slides.pdf

Apr 2026: