Keywords: Spatial data, simple features, raster data, time series, tidyverse


Working with spatial data in R goes back to the early days of R itself. For many spatial statistics and spatial analysis methods, R provides reference implementations. Spatial classes and methods in packages sp and raster, together with interfaces to GEOS and GDAL in rgeos, rgdal along with a few hundred packages depending on those have formed the cornerstone for many analysts for a long time.

Recent developments in the areas of data standardisation, web-based visualisation and computing, spatial databases, scalability, as well as R developments such as pipe-based workflows and the tidyverse have stimulated to rethink the way we handle spatial data in R. This has for instance resulted in package sf, a package that has been developed with support from the R consortium.

This workshop will illuminate old and the new ways of handling spatial data in R, will put some focus on handling simple features, and will discuss challenges ahead.

Learning objectives

After this workshop, participants should understand

Tutorial content

A tentative overview of the tutorial content is:


It is assumed that participants are familiar with R, and have some notion of spatial data. It is not required that they have a working knowledge of all the terms mentioned in this workshop description.

Potential attendees

Potential workshop attendees include data scientists and people working in a domain where they are faced with spatial data, e.g. ecologists, hydrologists, climate scientists, epidemiologists, social scientists, geographers, geoscientists, business intelligence analysts, and so on.

Instructor biography

Edzer Pebesma is full professor at the institute for geoinformatics of the university of Muenster. He is one of the authors of the book Applied Spatial Data Analysis with R, second edition, and is author and maintainer of a handful of packages found on his github page.

The material for this tutorial will also be developed on github.