Working with network data and statistical models using the R package MetricGraph
Alexandre De Bustamante Simas (KAUST)
Thursday 29th October 12:00-13:00
Maths 311B
Abstract
In this hands-on session, I will introduce MetricGraph, an R package for working with data and statistical models on metric graphs, such as road and river networks.
We will start by exploring the basic functionality of the package, including how to construct, visualize, and manipulate metric graph objects. I will demonstrate how to add observations and covariates to a graph, organize and group data, and extract, transform, and visualize information associated with the network.
We will then explore how to fit statistical models and perform prediction using INLA and inlabru, with examples involving Gaussian Whittle–Matérn fields. I will also illustrate more general applications, including log-Gaussian Cox processes and spatio-temporal models on networks.
The session will focus on practical examples in R, illustrating how MetricGraph can be used throughout a statistical analysis on networks. No previous experience with metric graphs is required, although some familiarity with R will be helpful.
Add to your calendar
Download event information as iCalendar file (only this event)