DM515 – Introduction to linear and integer programming

• Windows: GUSEK
• Linux and Mac: download GLPK, untar and execute:

  mkdir ~/usr

  ./configure –prefix=/home/yourid/usr

  make

  make install

  The exectuable is then ~/usr/bin/glpsol, or you can add the path in your bash config file:

  echo "export PATH=PATH:~/usr/bin" >> ~/.bashrc

  or tcsh config file (check echo $SHELL to know which shell you are using)

  echo "setenv PATH PATH:~/usr/bin" >> ~/.tcshrc

  As text editor you can use emacs that offers syntax highlighting, provided the extension of the model file is .mod.

The main reference is the documentation file gmpl.pdf (70 pages) from doc/ directory in the installation directory. Moreover check the files .mod in the directory examples/.

These might also be worth skimming through:

• The GNU Linear Programming Kit, Part 1: Introduction to linear optimization
• The GNU Linear Programming Kit, Part 2: Intermediate problems in linear programming
• The GNU Linear Programming Kit, Part 3: Advanced problems and elegant solutions

• To solve with GLPK solver:

  glpsol -m plan.mod -o plan.sol

• To solve with Gurobi:

  glpsol –check –model plan.mod –wlp plan.lp

  gurobi.sh

  gurobi> m=read("plan.lp")

  gurobi> m.optimize()

• To solve with SCIP from ZIBOPT

  glpsol –check –model plan.mod –wlp plan.lp

  scip -f plan.lp

• To solve with CPLEX

  glpsol –check –model plan.mod –wlp plan.lp

  cplex plan.lp > read plan.lp > optimize > help

Why GLPK?