Do what it says on the tin; make `mds2dd` work in dplyr pipelines, likely modifying it to accept one vector at a time