Online
Generalized linear mixed models are GLMs with random effects. This is a class of models allowing non-normal outcomes and dependencies between observations with applications in analysis of repeated observations, spatial data and genetics.
This lunch-to-lunch workshop is expected to be of interest especially to researchers and PhD students working with binary data applied on spatial data or genetics. It will give hands-on exercises using several packages in R starting with lme4. The advantages of using Bayesian methods such as the ones implemented in the R packages MCMCglmm and INLA will also be presented, as well as advanced extensions using the hglm package in R developed by the workshop organizer himself.
It is expected that the participants are acquainted/familiar with linear mixed models and GLMs. The workshop is focused on logistic regression with random effects.
Workshop leader: Lars Rönnegård
Date and time: 9:00- 12:00 on the 12-13 of June
Registration is required to participate in the workshop, sign up by clicking this link.
Day 1
• A quick recap on linear mixed models and GLMs. Show some nice applications with linear mixed models.
• Explain why it is difficult to fit GLMMs. Summarize methods to fit GLMMs; Laplace approximation, Gauss-Hermit Quadrature, and MCMC.
• Introduce the participants to the lme4 package
• Exercise with the lme4 package
• Introduce the participants to the MCMCglmm package
• Exercise with the MCMCglmm package
Day 2
• Discussion on how user-friendly the lme4 and MCMCglmm packages are.
• Introduction to the INLA package
• Exercise with INLA
• Introduction to the hglm package
• Exercise with hglm
• A summary and conclusions
The computer exercises will focus on logistic regression with random effects.