Comment by osipov

There is a big push for reproducibility in science. If you friend can package the process for building that graph in a Dockerfile, it is more likely that readers of her paper will be able to reproduce her results.