Introduction to Bayesian spatial statistics with inlabru in R

Prof Finn Lindgren will lead an introductory two-day course on the inlabru software on May 19th and 20th 2025

Event Overview

Bayesian Latent Gaussian Models (LGMs) are closely related to Generalized Additive Models (GAMs), offering Bayesian estimation and uncertainty quantification for spatial and spatio-temporal models. The INLA and inlabru R packages combine these Gaussian process models with numerical optimization and integration techniques, in a fast and flexible analysis toolkit. The taught part of the course will provide an overview of LGM theory and the INLA/inlabru methods and software, while the hands-on sessions will make sure the attendees will be ready to start doing spatial LGM modelling in R as soon as the course is over.

Registration

Please note that we will contact you separately for payment to confirm your place.

Fees

Fees include lunch and coffee on both days

Students: Free
University of Edinburgh Staff: £30
Higher Education and Research Institutions: £60
Industry: £200
Online only: £POA
Group discounts available: £POA

For POA please email DirectorCfS@ed.ac.uk for further details.

Topics covered

Basics of latent Gaussian process models in the Bayesian spatial statistics context. The principles of the INLA method for fast Baysian inference, and inlabru extensions for non-linear models. The inlabu package principles and interface. Building spatial and spatio-temporal models for point-referenced, spatially aggregated, and point pattern observations. Computing and assessing posterior predictions and visualisation. Diagnosing modelling problems.

Instructor

Prof Finn Lindgren

Finn Lindgren is Chair of Statistics in the School of Mathematics, University of Edinburgh.

His research focuses on spatial and spatio-temporal stochastic models, environmetrics, and computational methods and software. Among many others, he co-authored the influential paper “An Explicit Link Between Gaussian Fields and Gaussian Markov Random Fields: The Stochastic Partial Differential Equation Approach,” published in the Journal of the Royal Statistical Society: Series B. Professor Lindgren has contributed to the development of several R packages, including INLA for Bayesian latent Gaussian models and inlabru, a user-friendly interface for INLA with additional features

Learning outcomes

Understand the basic theory underpinning spatial latent Gaussian process models and Bayesian inference
Use inlabru to fit various spatial models to data, including point-referenced, aggregated, and point pattern data.
Be able to compute posterior predictions.
Assess and compare models

Daily timetable

To be announced soon

Target audience

Anyone with some statistics training who is aware of spatial data and the basics of additive models could benefit from attending. Fields where this may be most popular are: ecology, geosciences, ecology, epidemiology, public health, psychology, econometrics.

Assumed knowledge

Attendees should be comfortable with using R. They should understand linear and additive models, though this can be intuitive and doesn’t have to be mathematically rigorous. They do not need to have used INLA or inlabru before.