GLMAEDr

GLMAEDr is an R package for running the General Lake Model (GLM) coupled with the Aquatic EcoDynamics (AED) library. Rather than bundling pre-compiled executables (like GLM3r), GLMAEDr compiles GLM from source on your machine — avoiding code-signing restrictions on macOS and producing a binary that is tuned for your system.

System requirements

Install these before calling glm_install().

Platform	Command
macOS (Homebrew)	`brew install gcc netcdf gd`
Ubuntu / Debian	`sudo apt-get install gfortran libnetcdf-dev libgd-dev debhelper`
Windows	Install RTools for your R version. Note: Windows compilation support is experimental.

To check whether all build dependencies are present:

GLMAEDr::check_build_tools()

Installation

Install the package from GitHub with pak:

# install.packages("pak")
pak::pkg_install("FLARE-forecast/GLMAEDr")

Getting started

Step 1 — compile GLM

Run this once after installing the package. It clones the GLM-AED source repository and compiles the binary (~2 minutes):

library(GLMAEDr)
glm_install()

The binary is stored in a per-user data directory (tools::R_user_dir("GLMAEDr", "data")), so it survives package upgrades and requires no administrator privileges.

Confirm the installation succeeded:

glm_is_installed()   # TRUE
glm_version()        # prints the GLM version string

Step 2 — set up a simulation

Each simulation lives in its own folder and must contain a glm3.nml configuration file. Copy the bundled template as a starting point:

dir.create("my_lake")
file.copy(nml_template_path(), "my_lake/glm3.nml")

Edit glm3.nml to configure your lake geometry, meteorological inputs, and output settings.

Step 3 — run the simulation

run_glm("my_lake")

GLM writes its output to my_lake/output/output.nc by default. Verbose output is printed to the console; pass verbose = FALSE to suppress it.

Installing from AED_Tools

If you need the v4 development branch or have access to the private AED_Tools repository, use glm_install_aed_tools() instead:

# Public AED_Tools repo, v4alpha branch
glm_install_aed_tools(ref = "v4alpha")

# Private AED_Tools_Private repo (requires GitHub credentials)
glm_install_aed_tools(private = TRUE, ref = "v4alpha")

Registering a pre-built binary

If you already have a GLM executable (e.g. built manually or received from a colleague), register it directly without compiling from source:

glm_register_binary("~/Downloads/glm")

Upgrading GLM

Recompile with the latest source at any time:

glm_install(force = TRUE)

Pin a specific tag or branch:

glm_install(ref = "v3.9.107", force = TRUE)

Using GLM outside R

glm_path() returns the full path to the binary, useful for calling GLM from a shell script or other tool:

glm_path()
#> /Users/you/Library/Application Support/R/GLMAEDr/glm

Troubleshooting

Problem	Fix
`Missing build tools` error	Run `check_build_tools()` for platform-specific install instructions
Build fails on macOS	Ensure Homebrew `gcc` is on your PATH: `echo $PATH`
Build fails on Linux	Check that `libnetcdf-dev` is installed: `dpkg -l libnetcdf-dev`
`GLM is not installed` when calling `run_glm()`	Run `glm_install()` first
Want to try a different GLM version	`glm_install(ref = "v3.9.107", force = TRUE)`