Geocomputation with R — useR! 2026 tutorial

Materials for the hands-on tutorial Geocomputation with R at useR! 2026, by Jannes Muenchow, Robin Lovelace and Jakub Nowosad, authors of the book Geocomputation with R (2nd edition).

The slides are built with Quarto reveal.js (the modern successor to the xaringan decks used for the EGU 2019 version).

Structure

• slides/ — reveal.js decks (.qmd):
  • 01-intro.qmd — introduction, definitions and a brief spatial history
  • 02-crs.qmd — coordinate reference systems
  • 03-vector.qmd — vector data with the sf package
  • 04-raster.qmd — raster data with the terra package
  • 05-tmap.qmd — making maps with tmap
  • 06-qgisprocess.qmd — bridges to GIS software with qgisprocess (bonus)
• solutions/ — R scripts with exercise solutions
  • 02-crs-solution.R — CRS exercise solution
  • 03-vector-solution.R — vector exercise solution
  • 04-raster-solution.R — raster exercise solution
  • 05-tmap-solution.R — tmap exercise solution
• code/ — R scripts for the tmap code examples
• img/ — images used in the slides
• custom.scss — reveal.js theme
• references.bib — bibliography
• _quarto.yml — shared project configuration

Building the slides

The local R installation may not contain the geospatial stack, so the decks are rendered inside Docker using the ghcr.io/geocompx/latest image, which bundles R, RStudio, Quarto and all required packages (sf, terra, tmap, spData, leaflet, mapview, ...). Other variants (incl. the full QGIS stack, ghcr.io/geocompx/qgis) are documented at https://github.com/geocompx/docker.

Render a single deck:

docker run --rm -v "$PWD":/work -w /work ghcr.io/geocompx/latest \
  quarto render slides/01-intro.qmd

Render the whole project:

docker run --rm -v "$PWD":/work -w /work ghcr.io/geocompx/latest \
  quarto render

On Apple Silicon add --platform linux/amd64 (the images are amd64-only; OrbStack runs them under Rosetta).

VS Code Dev Container

A .devcontainer/ config is included. In VS Code, run Dev Containers: Reopen in Container to get R, RStudio, Quarto and all spatial packages inside the ghcr.io/geocompx/latest image, with the R and Quarto extensions pre-configured. Then use quarto preview slides/01-intro.qmd from the integrated terminal.

Plot pane troubleshooting notes are in .devcontainer/README.md.

Running RStudio (for attendees)

Attendees don't render slides — they run the exercise code in the RStudio IDE the image ships:

docker run -e PASSWORD=pw --rm -p 8787:8787 \
  ghcr.io/geocompx/latest

Open http://localhost:8787 and log in with rstudio / pw (add --platform linux/amd64 on Apple Silicon).

Local install

To develop locally instead, install R (≥ 4.3), Quarto and the packages listed in index.qmd, then run quarto render or quarto preview.