Skip to contents

Solve GMPL model using glpsol command line tool

Source: R/glpk_solver.R
solve_gmpl.Rd

Solves an optimization model by running GLPK's glpsol executable.

Usage

solve_gmpl(
  model = NULL,
  model_dir,
  method = "glpsol",
  verbose = TRUE,
  load_results = !is.null(model),
  glpsol_path = get_multimod_glpsol(),
  timeout = NULL
)

Arguments

model

Optional model object (for loading results back)

model_dir

Directory containing model files (model.mod, data.dat, etc.)

method

Retained for compatibility; only "glpsol" is supported. "glpkAPI" was retired when glpkAPI left CRAN - see drafts/glpkAPI-retired.R.

verbose

Logical; print progress messages (default: TRUE)

load_results

Logical; load solution CSV results back into model object (default: !is.null(model))

glpsol_path

Path to the glpsol executable; defaults to get_multimod_glpsol().

timeout

Numeric; timeout in seconds (default: NULL, no timeout)

Value

List with: model (if load_results=TRUE), success, objective, solve_time, status

Details

glpsol must be on the system: get_multimod_glpsol() resolves it from options("multimod.glpsol_path"), MULTIMOD_GLPSOL, the YAML config, the conda environment mm_setup_python() installs GLPK into, and finally the bare name "glpsol" on PATH.