在R？
中执行k-means后，是否有可能获得每个簇内的实际观测值例如，在我的分析之后，我有2个聚类，我想在每个聚类中找到确切的观察结果，是否可能？
答案 0 :(得分：4)
# random samples
x <- matrix(c(rnorm(30,10,2), rnorm(30,0,1)), nrow=12, byrow=T)
# clustering
clusters <- kmeans(x, 2)
# accessing cluster membership
clusters$cluster
[1] 1 1 1 1 1 1 2 2 2 2 2 2
# samples within cluster 1
c1 <- x[which(clusters$cluster == 1),]
# samples within cluster 2
c2 <- x[which(clusters$cluster == 2),]
# printing variables
x
[,1] [,2] [,3] [,4] [,5]
[1,] 10.8415151 9.3075438 9.443433171 13.5402818 7.0574904
[2,] 6.0721775 7.4570368 9.999411972 12.8186182 6.1697638
[3,] 11.3170525 10.9458832 7.576416396 12.7177707 6.7104535
[4,] 8.1377999 8.0558304 9.925363089 11.6547736 9.4911071
[5,] 11.6078294 8.7782984 8.619840508 12.2816048 9.4460169
[6,] 10.2972477 9.1498916 11.769122361 7.6224395 12.0658246
[7,] -0.9373027 -0.5051318 -0.530429758 -0.8200562 -0.0623147
[8,] -0.7257655 -1.1469400 -0.297539831 -0.0477345 -1.0278240
[9,] 0.7285393 -0.6621878 2.914976054 0.6390049 -0.5032553
[10,] 0.2672737 -0.6393167 -0.198287317 0.1430110 -2.2213365
[11,] -0.8679649 0.3354149 -0.003510304 0.6665495 0.6664689
[12,] 0.1731384 -1.8827645 0.270357961 0.3944154 1.3564678
c1
[,1] [,2] [,3] [,4] [,5]
[1,] 10.841515 9.307544 9.443433 13.540282 7.057490
[2,] 6.072177 7.457037 9.999412 12.818618 6.169764
[3,] 11.317053 10.945883 7.576416 12.717771 6.710454
[4,] 8.137800 8.055830 9.925363 11.654774 9.491107
[5,] 11.607829 8.778298 8.619841 12.281605 9.446017
[6,] 10.297248 9.149892 11.769122 7.622439 12.065825
c2
[,1] [,2] [,3] [,4] [,5]
[1,] -0.9373027 -0.5051318 -0.530429758 -0.8200562 -0.0623147
[2,] -0.7257655 -1.1469400 -0.297539831 -0.0477345 -1.0278240
[3,] 0.7285393 -0.6621878 2.914976054 0.6390049 -0.5032553
[4,] 0.2672737 -0.6393167 -0.198287317 0.1430110 -2.2213365
[5,] -0.8679649 0.3354149 -0.003510304 0.6665495 0.6664689
[6,] 0.1731384 -1.8827645 0.270357961 0.3944154 1.3564678