change process(), add nearestClusterIndex()

[charlesgriffiths]

process() wasn't working for my data set, it always put everything into the last cluster changed process() to be more in the spirit of iteration(), so each call to process() makes an improvement but the solution may not be a local minimum