Right now the pipeline defaults to do Louvain clustering. Maybe we can add an optional argument to do Leiden clustering if the user wants to do so.