Installing INLA in R on Windows

Working with Bayesian spatial models in R often brings its own set of challenges, especially when installing specialized packages like INLA. Recently, while preparing to model hotspots, I ran into exactly this issue on Windows. Here’s how I solved it and what INLA can do for your epidemiology projects.

The Challenge: Installing INLA on Windows

Started by trying the standard install command:

install.packages("INLA",
                 repos = "https://inla.r-inla-download.org/R/stable", dep = TRUE)

At first, everything looked fine. But then hit a roadblock:

• R tried to install the latest source version of INLA.

• Windows computers don’t have a compiler by default, so building from source failed.

• Error messages like:

ERROR: installing binary package failed

were appearing. This is common on Windows without Rtools.

The Solution: Installing the Binary Version

The fix was simple once I knew the trick. Force R to install the binary version of INLA. This avoids compilation entirely and works perfectly for most projects.

install.packages("INLA",
                 repos = "https://inla.r-inla-download.org/R/stable",
                 type = "binary")

• Works quickly

• Reliable

• Compatible with Windows

After this, INLA installed smoothly into our R library and was ready to use.

Checking the Installation

It is good practice to test that the package works:

library(INLA)

# Fit a tiny Poisson model as a test
data("kidney")
formula <- Count ~ 1 + f(Indiv, model="iid")
result <- inla(formula, family="poisson", data=kidney)

summary(result)

If you see the summary without errors, INLA is ready.

What INLA Does

INLA (Integrated Nested Laplace Approximation) is a powerful tool for Bayesian modeling, especially in spatial and spatiotemporal epidemiology. Common uses include:

• Estimating disease incidence with spatial and temporal variation

• Modeling environmental and socio-demographic predictors of disease

• Accounting for uncertainty in small or sparse datasets

• Producing posterior probability maps of disease risk

Other Useful Libraries

Alongside INLA, used several R packages to manage data and visualize results:

Package	Purpose
sf	Handling spatial data (points, polygons, shapefiles)
raster / terra	Environmental rasters like temperature, vegetation, elevation
ggplot2	Visualizing risk maps and model results
dplyr / data.table	Cleaning and manipulating data
tmap	Thematic maps for spatial data visualization
spdep	Spatial dependencies and neighborhood structures