You wouldn't have a dictionary when creating vectors from CSV (via CsvIterator). If u would like to see the documents that are part of cluster, try running the cluster output thru a seqdumper and that should give the document names (or points) that belong to a cluster.
You need to be working off of Mahout 0.9 or trunk to see the later working. On Friday, February 7, 2014 9:51 AM, "Allen, Ronald L." <allen...@ornl.gov> wrote: Hi, I've been able to get a CSV file into a sequence file of vectors readable by Mahout. I have ran mahout kmeans and it seems to work. But when I run mahout clusterdump, it does not work because I do not have a dictionary.file-0. Is there a way around this or a way to create this file myself? Thanks, Ronald
