用于分层聚类的聚类质心

Question

我使用hclust()对某些文本数据使用stringdist执行了层次式聚类。我在字符串之间得到了一个相异矩阵，并将其命名为distancemodels。

现在我正在尝试使用以下代码找到每个群集的中心：

dists = as.data.frame(distancemodels)
dists$ID = as.integer(rownames(dists))

# this adds the clusters information
dists = merge(dists,clusters[,c(1,4)])

#k = number of clusters
meds = as.vector(1:k)

#This for loop is throwing the following error: Error in colMeans(dists[dists$cluster == i, as.character(dists$ID[dists$cluster ==  :'x' must be an array of at least two dimensions
for(i in 1:k){meds[i] = as.integer(names(colMeans(dists[dists$cluster ==      i,as.character(dists$ID[dists$cluster == i])])[unname(which(colMeans(dists[dists$cluster == i,as.character(dists$ID[dists$cluster == i])])==min(colMeans(dists[dists$cluster == i,as.character(dists$ID[dists$cluster == i])]))))]))[1]}

medians = as.data.frame(unlist(t(t(meds))))
medians$cluster = rownames(medians)

由于我无法从互联网上找到有关如何为hclust找到群集质心的任何帮助，这是我写的。请让我知道我哪里出错了。我是R的新手。