Handling and Analyzing Spatio-temporal Data in R

Tutorial: Handling and Analyzing Spatio-temporal Data in R

Roger Bivand, Norwegian School of Economics and Business Administration, Norway. (Instructor)
Edzer Pebesma, University of Münster, Germany. (Co-author)

Overview

This hands-on tutorial will permit participants to learn how to begin representing and using spatio-temporal data in R. This topic is emerging as one of the more challenging and promising ones in a wide range of disciplines, especially as data collection procedures become faster and more efficient. Getting an early view of what is happening should be potentially useful to many applied researchers from fields including but not limited to environmental sciences, epidemiology, ecology, social sciences, geography, and economics.

Goals

After the tutorial, participants will

have an overview of how spatio-temporal data may be represented in R;
know which contributed packages may be used with base functionality to handle spatio-temporal data;
be able to visualize some types of spatio-temporal data;
be able to analyse some types of spatio-temporal data.

Outline

R provides rich environments for handling and analyzing spatial or time series data, but spatio-temporal data is typically dealt with on an ad-hoc basis, with methods that deal with space and/or time implicitly as factors, or numerically without proper space or time reference frames. This tutorial will review existing approaches, and discuss examples from socio-economic data in case of constant or changing boundaries, trajectory data from ecology, data from moving sensors, spatio-temporal point pattern data, and spatio-temporal interpolation. It will discuss the benefits and limitations of ad hoc approaches, and point out the potential of and challenges to an integrated, referenced spatio-temporal approache which is e.g. pioneered in the spacetime package.

Prerequisites

Participants should have some knowledge of how spatial and/or temporal data are represented in R, or at least should have an interest in working with data of this kind, and feel comfortable with using R. Participants will need a laptop, with current R installed, together with a list of contributed packages to be provided and updated in the workshop materials below. Downloading the tutorial datasets and scripts from the same site before attending will make it possible to follow the tutorial hands-on.

Intended Audience

R users working with, or about to start working with spatio-temporal data.

Workshop Materials

Workshop materials will be published on this site well before the tutorial begins, and will include copies of slides, lists of packages, scripts, and data sets.