On Dec 2, 2011, at 4:47 PM, Gary Snider wrote: > Displaykmeans uses a DenseVector of 2D points. Very easy to visualize due > to x,y coordinates. > > The reuters dataset uses RandomAccessSparseVectors and produces a sequence > file with the two types (org.apache.hadoop.io.Text & > org.apache.mahout.clustering.kmeans.Cluster) > > If you find a way to visualize the reuters clusters PLEASE post it.
Have a look at MAHOUT-899. Also see the recent thread between me and Ted on visualizing clustering. Ted posts a nice little R script that can be used to visualize things. On MAHOUT-899, I've put up some mods to the ClusterDumper that make it easy to display things in Gephi. > > > > -----Original Message----- > From: syed kather [mailto:[email protected]] > Sent: Friday, December 02, 2011 10:00 AM > To: [email protected] > Subject: DisplayKMean > > > > Team , > > > > Can you help me how to do execute Displaykmeans . As i seen there are two > path is needed on is "sample" and "output" . If so can you help me what are > path should be passed for this .. Consider the examples ret > > > > > > Thanks and Regards, > > S SYED ABDUL KATHER > > 9731841519 -------------------------------------------- Grant Ingersoll http://www.lucidimagination.com
