• Cvxpy_polymatroid contains the problem setup and definitions
• Cvxpy_experimentation contains a bunch of (often messy and ad-hoc) tests run on the data. Should it be a Jupyter notebook? Ideally, probably yes.
• Dense_Cvxpy contains a sanity check, running the problem on statistics describing a dense matrix, to make sure we recover the dense bound.