r - 选择两列之间具有最大值的列值

Question

我有一个名为

的数据框

tt <- structure(list(IID1 = c("035-0028-0151526", "036-1002-0282844", 
"950-0001-0081329", "951-1000-0084436", "954-0009-0337439", "954-0011-0379455", 
"955-0017-0354237", "6011039", "6011039", "6011039", "6011039", 
"6205526", "A1858", "A2159"), IID2 = c("035-0028-0157134", "036-1002-0305628", 
"950-0001-0100509", "951-1000-0084537", "954-0009-0337641", "954-0011-0379657", 
"955-0017-0363137", "6205526", "6215777", "A1864", "A2267", "A1864", 
"A1864", "A2267"), nIID1 = c(1, 1, 1, 1, 1, 1, 1, 4, 4, 4, 4, 
1, 1, 1), nIID2 = c(1, 1, 1, 1, 1, 1, 1, 1, 1, 3, 2, 3, 3, 2)), row.names = c(NA, 
-14L), class = c("data.table", "data.frame"), .internal.selfref = <pointer: 0x38b6d80>)

我想从 IID1 和 IID2 中提取值并将其放入 selected 列中。如果nIID1 == nIID2，我想要来自IID1的值，否则，我想要来自具有最大IID的列的值。例如，如果 nIID1 的值高于 nIID2，那么我需要 IID1 的值。有人可以指导我解决方案吗？谢谢!

结果应如下所示。

               IID1             IID2 nIID1 nIID2         selected
 1: 035-0028-0151526 035-0028-0157134     1     1 035-0028-0151526
 2: 036-1002-0282844 036-1002-0305628     1     1 036-1002-0282844
 3: 950-0001-0081329 950-0001-0100509     1     1 950-0001-0081329
 4: 951-1000-0084436 951-1000-0084537     1     1 951-1000-0084436
 5: 954-0009-0337439 954-0009-0337641     1     1 954-0009-0337439
 6: 954-0011-0379455 954-0011-0379657     1     1 954-0011-0379455
 7: 955-0017-0354237 955-0017-0363137     1     1 955-0017-0354237
 8:          6011039          6205526     4     1          6011039
 9:          6011039          6215777     4     1          6011039
10:          6011039            A1864     4     3          6011039
11:          6011039            A2267     4     2          6011039
12:          6205526            A1864     1     3            A1864
13:            A1858            A1864     1     3            A1864
14:            A2159            A2267     1     2            A2267

Answer 1

我们可以通过max.col获取列索引，通过cbind获取行索引(.I)，从cbinded IID1、IID2 并将 (:=) 分配给“选定”

tt[, selected := cbind(IID1, IID2)[cbind(.I, max.col(.SD, "first"))],
      .SDcols = nIID1:nIID2]

-输出

> tt
                IID1             IID2 nIID1 nIID2         selected
 1: 035-0028-0151526 035-0028-0157134     1     1 035-0028-0151526
 2: 036-1002-0282844 036-1002-0305628     1     1 036-1002-0282844
 3: 950-0001-0081329 950-0001-0100509     1     1 950-0001-0081329
 4: 951-1000-0084436 951-1000-0084537     1     1 951-1000-0084436
 5: 954-0009-0337439 954-0009-0337641     1     1 954-0009-0337439
 6: 954-0011-0379455 954-0011-0379657     1     1 954-0011-0379455
 7: 955-0017-0354237 955-0017-0363137     1     1 955-0017-0354237
 8:          6011039          6205526     4     1          6011039
 9:          6011039          6215777     4     1          6011039
10:          6011039            A1864     4     3          6011039
11:          6011039            A2267     4     2          6011039
12:          6205526            A1864     1     3            A1864
13:            A1858            A1864     1     3            A1864
14:            A2159            A2267     1     2            A2267

---

或者另一个选项是fcase

tt[, selected := fcase(nIID1 >= nIID2, IID1, nIID1 < nIID2, IID2)]

编辑:按照@Marco_CH的建议