I'm trying to use Spicker to cluster a set of 2000 structures. I've done two experiments, changing only the order of structures in the input file rep.tra.
I've noticed that the resulting clusters are different in their population in the two cases (the last three of the resulting eight clusters have a different number of structures in the two experiments ).
According to the clustering algorithm, I can't understand how order of input structures affects the output
clusters.
Thank you in advance and all the best.