from sklearn.cluster import KMeans
from sklearn.datasets import make_blobs
import matplotlib.pyplot as plt
from sklearn.metrics import silhouette_score #所有样本轮廓系数均值
from sklearn.metrics import silhouette_samples #每一个样本的轮廓系数#自己创建数据集
X,y=make_blobs(n_samples=500,n_features=2,centers=4,random_state=1)#规定randomstate是让数据稳定#基于这个分布进行聚类
#kmeans 并不需要建立模型或预测结果 只需要fit就可以得到聚类结果
#kmeans 有接口predict和fit_predict 表示学习数据X并对X的类别进行预测
#但得到结果后一般不调用predict 直接fit之后调用熟悉labels_一模一样
#一般当数据量大时，才使用predict
n_cluster=3
cluster=KMeans(n_clusters=n_cluster,random_state=1).fit(X)
#查看聚好的类别 每个样本对应的类
y_pred=cluster.labels_
#查看质心
centroid=cluster.cluster_centers_
#查看总距离平方和
inertia=cluster.inertia_
# print(centroid)
# print(inertia)color = ["red","pink","orange"]
label=['cluster1','cluster2','cluster3']
fig,ax1=plt.subplots(1)for i in range(len(color)):ax1.scatter(X[y_pred==i,0],X[y_pred==i,1],marker='o',#点的形状s=8,#点的大小label=label[i],c=color[i])#画出质心
ax1.scatter(centroid[:,0],centroid[:,1],marker='x',s=15,c='black')plt.legend(loc=0)
# plt.show()#查看轮廓系数
print(silhouette_score(X,y_pred))
print(silhouette_samples(X,y_pred))

分析n_cluster=4时，分簇效果

from sklearn.cluster import KMeans
from sklearn.datasets import make_blobs
import matplotlib.pyplot as plt
import matplotlib.cm as cm
import numpy as np
from sklearn.metrics import silhouette_score #所有样本轮廓系数均值
from sklearn.metrics import silhouette_samples #每一个样本的轮廓系数"""基于轮廓系数调整n_cluster求每一个聚出来的类的轮廓系数各个类之间轮廓系数的对比聚类后图像的分布情况绘制学习曲线
"""
#创建数据
X,y=make_blobs(n_samples=500,n_features=2,centers=4,random_state=1)#规定randomstate是让数据稳定n_clusters=4
fig,(ax1,ax2)=plt.subplots(1,2)
fig.set_size_inches(18,7)#设置画布尺寸
#第一个图是轮廓系数图像（横向条形图） 横坐标为轮廓系数 纵坐标为每一个样本
ax1.set_xlim([-0.1,1])#设置轮廓系数范围
ax1.set_ylim([0,X.shape[0]+(n_clusters+1)*10])#加上后面的数是为了显示时有一定的间隔
clusterer=KMeans(n_clusters=n_clusters,random_state=10).fit(X)
cluster_labels=clusterer.labels_ #聚类后每一个样本对应的类silhouette_avg = silhouette_score(X, cluster_labels)#平均轮廓系数
print("For n_clusters ={} The average silhouette_score is :{}".format(n_clusters, silhouette_avg))
#每个样本的轮廓系数 用作横坐标
sample_silhouette_values = silhouette_samples(X, cluster_labels)
y_lower=10 #设置y轴初始取值 避免贴着x轴开始画图#对每一个簇进行循环
for i in range(n_clusters):#从每个样本的轮廓系数取出第i个簇的轮廓系数ith_cluster_silhouette_values = sample_silhouette_values[cluster_labels == i]ith_cluster_silhouette_values.sort()size_cluster_i=ith_cluster_silhouette_values.shape[0] #查看这个簇里有多少样本y_upper=y_lower+size_cluster_i    #这个簇在y轴的上界color=cm.nipy_spectral(float(i)/n_clusters) #输入任意一个浮点数 会返回该浮点数对应的函数 保证不同簇的颜色不同#fill_between是一个让范围中的柱状图都统一颜色的函数#fill_betweenx的范围在纵坐标上 fill_betweeny的范围在横坐标上#fill_betweenx 纵坐标下限 纵坐标上限 x取值 颜色ax1.fill_betweenx(np.arange(y_lower,y_upper),ith_cluster_silhouette_values,facecolor=color,alpha=0.7)#为每个簇的轮廓系数写上编号 并让编号显示在中间位置ax1.text(-0.05,y_lower + 0.5 * size_cluster_i, str(i))#更新y_lower 保证每一个簇直接存在间隔y_lower = y_upper + 10ax1.set_title("The silhouette plot for the various clusters.")
ax1.set_xlabel("The silhouette coefficient values")
ax1.set_ylabel("Cluster label")#把轮廓系数均值用虚线形式放入图中
ax1.axvline(x=silhouette_avg, color="red", linestyle="--")
ax1.set_yticks([])
ax1.set_xticks([-0.1, 0, 0.2, 0.4, 0.6, 0.8, 1])#画第二个图
colors = cm.nipy_spectral(cluster_labels.astype(float) / n_clusters)#利用分簇结果取n_clusters个颜色
centers = clusterer.cluster_centers_
ax2.scatter(X[:, 0], X[:, 1],marker='o',s=8,c=colors)#画分簇图
ax2.scatter(centers[:, 0], centers[:, 1], marker='x',c="red", alpha=1, s=200) #画质心
ax2.set_title("The visualization of the clustered data.")
ax2.set_xlabel("Feature space for the 1st feature")
ax2.set_ylabel("Feature space for the 2nd feature")#设置整张图片标题
plt.suptitle(("Silhouette analysis for KMeans clustering on sample data ""with n_clusters = {}".format(n_clusters)),fontsize=14, fontweight='bold')plt.show()

包装成循环之后查看n_cluster取不同值下的轮廓系数，并以图片形式显示。

from sklearn.cluster import KMeans
from sklearn.datasets import make_blobs
from sklearn.metrics import silhouette_samples, silhouette_score
import matplotlib.pyplot as plt
import matplotlib.cm as cm
import numpy as np#自己创建数据集
X,y=make_blobs(n_samples=500,n_features=2,centers=4,random_state=1)#规定randomstate是让数据稳定for n_clusters in [2, 3, 4, 5, 6, 7]:n_clusters = n_clustersfig, (ax1, ax2) = plt.subplots(1, 2)fig.set_size_inches(18, 7)ax1.set_xlim([-0.1, 1])ax1.set_ylim([0, X.shape[0] + (n_clusters + 1) * 10])clusterer = KMeans(n_clusters=n_clusters, random_state=10).fit(X)cluster_labels = clusterer.labels_silhouette_avg = silhouette_score(X, cluster_labels)print("For n_clusters =", n_clusters,"The average silhouette_score is :", silhouette_avg)sample_silhouette_values = silhouette_samples(X, cluster_labels)y_lower = 10for i in range(n_clusters):ith_cluster_silhouette_values = sample_silhouette_values[cluster_labels == i]ith_cluster_silhouette_values.sort()size_cluster_i = ith_cluster_silhouette_values.shape[0]y_upper = y_lower + size_cluster_icolor = cm.nipy_spectral(float(i) / n_clusters)ax1.fill_betweenx(np.arange(y_lower, y_upper), ith_cluster_silhouette_values, facecolor=color, alpha=0.7)ax1.text(-0.05, y_lower + 0.5 * size_cluster_i, str(i))y_lower = y_upper + 10ax1.set_title("The silhouette plot for the various clusters.")ax1.set_xlabel("The silhouette coefficient values")ax1.set_ylabel("Cluster label")ax1.axvline(x=silhouette_avg, color="red", linestyle="--")ax1.set_yticks([])ax1.set_xticks([-0.1, 0, 0.2, 0.4, 0.6, 0.8, 1])colors = cm.nipy_spectral(cluster_labels.astype(float) / n_clusters)ax2.scatter(X[:, 0], X[:, 1], marker='o', s=8, c=colors)centers = clusterer.cluster_centers_# Draw white circles at cluster centersax2.scatter(centers[:, 0], centers[:, 1], marker='x',c="red", alpha=1, s=200)ax2.set_title("The visualization of the clustered data.")ax2.set_xlabel("Feature space for the 1st feature")ax2.set_ylabel("Feature space for the 2nd feature")plt.suptitle(("Silhouette analysis for KMeans clustering on sample data ""with n_clusters = %d" % n_clusters),fontsize=14, fontweight='bold')
plt.show()

2.2 重要参数init & random_state & n_init 初始化质心