import numpy as np from scipy.cluster.hierarchy import dendrogram import matplotlib.pyplot as plt index_m0,index_m1 = 0,0 lk = [] Y = np.array([1,4,9,16,25,36,49,64,81]) mod = Y.copy() Mean = Y.mean() Y = Y - Mean Y = abs(Y) dic = {} for i in Y: dic[i] = 1 def fun(Y,Mean):#通过距离计算,输出最接近簇中心的两个点的索引 Y0 = Y.copy() index_m0, index_m1 = 0,0 index_m0 = list(Y0).index(min(Y0)) Y0[index_m0] = 1e6 index_m1 = list(Y0).index(min(Y0)) return index_m0, index_m1 while dic[Y[-1]] != 9: #当集群点的数量小于节点数,循环直到所有点加入集群并停止 index_m0,index_m1 = fun(Y,Mean) Y = list(Y) Y.append(Y[index_m0]+Y[index_m1]) dic[Y[-1]] = dic[Y[index_m0]] + dic[Y[index_m1]] #在两个连续的聚类之后,添加距离以获得后续聚类的新点 Y[index_m0] = 1e6 #将最近的两个点赋值为1e6,通过定义距离为无穷大来防止后续的重聚。 Y[index_m1] = 1e6 Y = np.array(Y) lk.append([index_m0,index_m1,Y[-1],dic[Y[-1]]]) #生成格式为[点,点,簇中心距离,簇点总数],便于后续可视化。 lk = np.array(lk) dendrogram(lk) plt.show()