在做heatmap聚类之后，常常因为一些分析，需要获取聚类后的矩阵。这里分享一种刚刚学习到的方法。
该方法基于heatmap.2的hierarchical clustering聚类。
首先产生随机数据作为演示,生成的数据如下(因为是随机生成，所以数字会不一样)
> before <- matrix(rnorm(50), 10, 5, dimnames=list(paste("g", 1:10, sep=""),paste("t", 1:5, sep="")))
> before
t1 t2 t3 t4 t5
g1 -1.89937785 -0.2112492 0.80475998 -0.1437429 -0.88203601
g2 0.75889366 1.5513556 -1.08355333 -1.7635301 -0.65163564
g3 0.53893684 -0.3485060 -0.86417609 0.6294090 0.33726749
g4 1.42944861 1.1869920 -0.50002826 0.8134602 -2.11838364
g5 0.25707340 0.8692485 -0.47245735 2.2251189 -0.05577767
g6 -0.36162410 -1.7610345 -0.04321973 -0.8802015 0.76510606
g7 0.95574696 -1.4834578 1.82203982 -1.1817150 -0.05954665
g8 0.03605131 -0.2152805 -0.05155960 2.1196481 0.79576233
g9 -2.39660463 1.1517748 0.38779793 -1.2362860 -0.83512569
g10 0.99780002 1.3384064 -1.45549251 -0.7208040 -1.65369287聚类：
> library(gplots) > test <- heatmap.2(before)
获取聚类之后的矩阵
> after = before[rev(test$rowInd), test$colInd]
> after
t5 t3 t4 t1 t2
g5 -0.05577767 -0.47245735 2.2251189 0.25707340 0.8692485
g8 0.79576233 -0.05155960 2.1196481 0.03605131 -0.2152805
g3 0.33726749 -0.86417609 0.6294090 0.53893684 -0.3485060
g4 -2.11838364 -0.50002826 0.8134602 1.42944861 1.1869920
g2 -0.65163564 -1.08355333 -1.7635301 0.75889366 1.5513556
g10 -1.65369287 -1.45549251 -0.7208040 0.99780002 1.3384064
g7 -0.05954665 1.82203982 -1.1817150 0.95574696 -1.4834578
g6 0.76510606 -0.04321973 -0.8802015 -0.36162410 -1.7610345
g1 -0.88203601 0.80475998 -0.1437429 -1.89937785 -0.2112492
g9 -0.83512569 0.38779793 -1.2362860 -2.39660463 1.1517748