In this post we look at why the bipartite matching problem returns an optimal integer solution even if we relax the integrality constraint.