现在要做聚类分析，根据文献，应该是
先做hierarchial clustering.从而确定出最佳的类别数。
然后第二步进行K-means的聚类，但要求将确定出的“clustering center”作为种子来进行这个聚类。
原文“two-step approach was used. First, hierarchical clustering was done to identify the number of clusters. Then, the cluster centers identified were used as starting seeds in a K-means nonhierarchical clustering procedure.”

问题在于，不知道这个“clustering center”是怎么得出的，是通过别的分析得出，还是hierarchial clustering分析中本身就能产生？求指教。谢谢！