cluster-tests Aim: use existing functions to cluster large vegetation datasets where some of the relevés can be a priori assigned to groups