afni.nimh.nih.gov

Hi Ziad,
I have another questions about 3dkmeans. For the -r option, the help says "-r number For k-means clustering, the number of times the k-means clustering algorithm is run (default: 0 with -clust_init, 1 otherwise)".
So if I use "-r 5", my understanding is that the k-means algorithm will run 5 times with 5 different sets of initial centers. Is my understanding correct? If yes, then how are these 5 clustering results combined to produce the final results?