(This is the Apr 10th version. It should all work, but the docker image might be updated until end of April. The changes will be minor, though.)

The docker image you should use is provided as a tar file on the this webpage. Note, that the image is approx. 5GB in size. The specific docker file does not come with a license file for the ILP solver Gurobi, which you can easily get via their webpage. Note, that you need a WLS license from Gurobi in order to use it inside docker. Alternatively, you can use the IBM CPLEX solver by setting the environment variable using export MOD_ILP_SOLVER="CPLEX" inside docker or config.ilp.solver="CPLEX" inside a python script.

mød using docker

First, you should install docker on your personal machine (see here ). Make sure you can use docker (e.g. test with the command docker run hello-world).

If this works, copy the mød image provided above to your local machine. You can then easily load a docker image with the command docker load -i mod-priv-0.17.1.14-nogurobi.lic.tar. This will take a while but has to be done only once. After the image was loaded you can start a container as explained on the mød installation webpage. I personally (on my MacBook) use the command docker run -it --hostname docker --rm -v $(pwd):/workdir -u 0:100 mod-priv:0.17.1.14-nogurobi.lic. When you run mød in the docker container, you are usually interested in the output generated in the summary directory, which is also accessible from the host system (in the directory where you started the docker container).

Note, that changes within the docker container are not persistent, but you can make persistent changes by committing changes to the container. See any docker introductory material in case you have the need for that.