如何在R中使用WeightedCluster::wcKMedoids为heatmap或heatmap.2提供群集?
TL;DR:如何使用如何在R中使用WeightedCluster::wcKMedoids为heatmap或heatmap.2提供群集?,r,cluster-analysis,heatmap,R,Cluster Analysis,Heatmap,TL;DR:如何使用WeightedCluster库(特别是wcKMedoids()方法)作为heatmap、heatmap.2或类似工具的输入,为其提供聚类信息 我们正在从R中的一些二进制数据(是/否值,表示为1和0)创建热图,并且需要为基于列的聚类调整一些行的权重 (它们从多项选择类别生成到多个二进制是/否值行,因此被过度表示) 我找到了这个库,它可以使用权重进行聚类 现在的问题是如何使用此库(特别是wcKMedoids()方法)作为heatmap、heatmap.2或类似方法的输入 我尝
WeightedCluster
库(特别是wcKMedoids()
方法)作为heatmap
、heatmap.2
或类似工具的输入,为其提供聚类信息
我们正在从R中的一些二进制数据(是/否值,表示为1和0)创建热图,并且需要为基于列的聚类调整一些行的权重 (它们从多项选择类别生成到多个二进制是/否值行,因此被过度表示) 我找到了这个库,它可以使用权重进行聚类 现在的问题是如何使用此库(特别是
wcKMedoids()
方法)作为heatmap
、heatmap.2
或类似方法的输入
我尝试了以下代码,结果出现以下错误消息:
library(gplots)
library(WeightedCluster)
dataset <- "
F,T1,T2,T3,T4,T5,T6,T7,T8
A,1,1,0,1,1,1,1,1
B,1,0,1,0,1,0,1,1
C,1,1,1,1,1,1,1,0
D,1,1,1,0,1,1,1,0
E,0,1,0,0,1,0,1,0
F,0,0,1,0,0,0,0,0
G,1,1,1,0,1,1,1,1
H,1,1,0,0,0,0,0,0
I,1,0,1,0,0,1,0,0
J,1,1,1,0,0,0,0,1
K,1,0,0,0,1,1,1,1
L,1,1,1,0,1,1,1,1
M,0,1,1,1,1,1,1,1
N,1,1,1,0,1,1,1,1"
fakefile <- textConnection(dataset)
d <- read.csv(fakefile, header=T, row.names = 1)
weights <- c(1,1,1,1,1,1,1,1,1,1,1,1,1,1)
distf <- function(x) dist(x, method="binary")
wclustf <- function(x) wcKMedoids(distf(x),
k=8,
weights=weights,
npass = 1,
initialclust=NULL,
method="PAMonce",
cluster.only = FALSE,
debuglevel=0)
cluster_colors <- colorRampPalette(c("red", "green"))(256);
heatmap(as.matrix(d),
col=cluster_colors,
distfun = distf,
hclustfun = wclustf,
keep.dendro = F,
margins=c(10,16),
scale="none")
显然,wcKMedoids
并不能替代R的hclust
,但是有人对如何解决这个问题有一些建议吗
更新:到目前为止,我取得的微小进展表明,我应该实现一个方法
作为.dendrogram.kmedoids
,该方法产生与hclust(dist(x))
类似的输出。(可以使用dput
详细检查其输出:dput(hclust(dist(x)))
)。非常欢迎想法和建议。这是不可能做到的。K-Medoid聚类是一种划分方法,而不是分层方法。Dendogram仅对分层聚类算法有意义。如果您可以使用更简单的解决方案,只需将权重乘以原始矩阵,以这种方式赋予它们更大的权重。我不是100%确定这是统计上正确的方法,但取决于你想要实现什么,它可能会起作用
# Create the dataset
dataset <- matrix(
dimnames = list(LETTERS[seq( from = 1, to = 14 )], c("T1","T2","T3","T4","T5","T6","T7","T8")),
data = c(1,1,0,1,1,1,1,1,
1,0,1,0,1,0,1,1,
1,1,1,1,1,1,1,0,
1,1,1,0,1,1,1,0,
0,1,0,0,1,0,1,0,
0,0,1,0,0,0,0,0,
1,1,1,0,1,1,1,1,
1,1,0,0,0,0,0,0,
1,0,1,0,0,1,0,0,
1,1,1,0,0,0,0,1,
1,0,0,0,1,1,1,1,
1,1,1,0,1,1,1,1,
0,1,1,1,1,1,1,1,
1,1,1,0,1,1,1,1),
ncol=8,
nrow=14)
# Assign weights to the different columns
col.weights <- c(2,3,1,1,1,1,1,1)
# Transform the original matrix with the weights
# you want to assign to each column.
create.weights.matrix <- function(weights, rows) {
sapply(weights, function(x){rep(x, rows)})
}
weights.matrix <- create.weights.matrix(col.weights, nrow(dataset))
d.weighted <- weights.matrix * dataset
# Create the plot
cluster_colors <- colorRampPalette(c("red", "green"))(256);
heatmap(as.matrix(d.weighted),
col=cluster_colors,
keep.dendro = F,
margins=c(10,16),
scale="none")
#创建数据集
dataset我投票结束这个问题,因为它是关于如何在没有可复制示例的情况下使用R的。@对此我很抱歉(对R有点陌生,以及如何做),使代码示例现在完全独立且可复制!
# Create the dataset
dataset <- matrix(
dimnames = list(LETTERS[seq( from = 1, to = 14 )], c("T1","T2","T3","T4","T5","T6","T7","T8")),
data = c(1,1,0,1,1,1,1,1,
1,0,1,0,1,0,1,1,
1,1,1,1,1,1,1,0,
1,1,1,0,1,1,1,0,
0,1,0,0,1,0,1,0,
0,0,1,0,0,0,0,0,
1,1,1,0,1,1,1,1,
1,1,0,0,0,0,0,0,
1,0,1,0,0,1,0,0,
1,1,1,0,0,0,0,1,
1,0,0,0,1,1,1,1,
1,1,1,0,1,1,1,1,
0,1,1,1,1,1,1,1,
1,1,1,0,1,1,1,1),
ncol=8,
nrow=14)
# Assign weights to the different columns
col.weights <- c(2,3,1,1,1,1,1,1)
# Transform the original matrix with the weights
# you want to assign to each column.
create.weights.matrix <- function(weights, rows) {
sapply(weights, function(x){rep(x, rows)})
}
weights.matrix <- create.weights.matrix(col.weights, nrow(dataset))
d.weighted <- weights.matrix * dataset
# Create the plot
cluster_colors <- colorRampPalette(c("red", "green"))(256);
heatmap(as.matrix(d.weighted),
col=cluster_colors,
keep.dendro = F,
margins=c(10,16),
scale="none")