r对于一个表，重新对每一行进行独立排序

Question

目前，我有一个类似以下的csv文件

ID      grade-1         grade-2         grade-3
 1    0.004461027   0.002740424 0.002955164
 2    0.055690775   0.045791653 0.17440305
 3    0.048901623   0.042439538 0.027306325
 4    0.20013265    0.0637944   0.081362503

我将表格读取为

test.matrix<-data.frame(read.table("test.csv",sep=",",header=T))

我想生成一个新的表，其中每一行都经过排序

ID      highest grade           the second grade           the third grade
1   grade-1:0.004461027 grade-3:0.002955164        grade-2:0.002740424  
2   grade-3:0.17440305      grade-1:0.055690775    grade-2:0.045791653  
3   grade-1:0.048901623 grade-2:0.042439538    grade-3:0.027306325
4   grade-1:0.20013265  grade-3:0.081362503        grade-2:0.0637944    

如何对每一行进行排序？为了生成输出，如何将字符(例如grade-1(和数值(例如0.004461027(放在单个条目中，例如grade-1:0.004461027？

Answer 1

也许：

 res <- t( apply( dfrm[ 2:4], 1, 
                    function(row) paste0("grade-", 1:3, ":", rev(sort(row) ) ) ) )

R按列顺序返回矩阵结果，因此当函数应用于行时，需要转置结果以获得序数"形状"。要将ID值放回，请cbind到ID:

 cbind(dfrm[, "ID", drop=FALSE], res)

我把drop等于FALSE放在那里，以保持第一个参数的dataframe类，这样结果将是data.frame。否则，res对象是矩阵，dfrm[，"ID"]或dfrm$ID将是向量，cbind结果将是矩阵。

Answer 2

t(apply(DF,1,function(x) {
  temp <- sort(x[-1],decreasing=TRUE)
  res <- c(x[1],paste(names(temp),temp,sep=": "))
  names(res) <- c("ID",      "highest grade",           "the second grade",           "the third grade")
  res
                        }))
     ID  highest grade          the second grade       the third grade       
[1,] "1" "grade.1: 0.004461027" "grade.3: 0.002955164" "grade.2: 0.002740424"
[2,] "2" "grade.3: 0.17440305"  "grade.1: 0.055690775" "grade.2: 0.045791653"
[3,] "3" "grade.1: 0.048901623" "grade.2: 0.042439538" "grade.3: 0.027306325"
[4,] "4" "grade.1: 0.20013265"  "grade.3: 0.081362503" "grade.2: 0.0637944"