Hi, The wiki says the default method for cluster.split is furthest neighbor while the default for cluster is average neighbor…are they really different defaults?
Thanks!
Hi, The wiki says the default method for cluster.split is furthest neighbor while the default for cluster is average neighbor…are they really different defaults?
Thanks!
Sorry - we missed this in the documentation. Average neighbor is the defualt for all clustering algorithms. Yes, there is a big difference - see Westcott & Schloss AEM 2012 for details.