我希望以距离矩阵的形式拥有相同大小的3个聚类,总共1000个观测值。以下是为简单起见而进行的5次观察的样本数据。
1 2 3 4 5
1 0 302.0366182 1802.347771 1929.003625 4152.862153
2 302.0366182 0 1502.097742 1629.91035 3966.183353
3 1802.347771 1502.097742 0 137.5130532 3439.08972
4 1929.003625 1629.91035 137.5130532 0 3470.441858
5 4152.862153 3966.183353 3439.08972 3470.441858 0
A <- as.dist(as(Clustering, "matrix"))
Clustering <- scale(Clustering) # standardize variables
# Ward Hierarchical Clustering
d <- dist(Clustering, method*emphasized text* = "euclidean") # for a distance matrix
fit <- hclust(d, method="ward.D2")
plot(fit) # display dendogram
cutree(fit, k=3) # cut tree into 3 clusters
# draw dendogram with red borders around the 3 clusters
rect.hclust(fit, k=3, border="red")
如何获得相同大小的群集?