在 R 中，如何将行中的第 n 个最大数和 output 结果返回到新列，每行重复？

Question

Let's say I am trying to select the 2nd highest value row-wise from a specified vector of columns (in this example, from columns 1 through 4). I then want the result (ideally, the actual number) to be sent to a new column, let's call it Second . How might I do this?

Current code using the Rfast package (the documenatation of which I can't decipher):

df$Second <- Rfast::rownth(as.matrix(df[,c(1:4)]), elems=3)

Here are some example data:

df <- structure(list(A = c(-0.113802816901408, -0.613802816901408, 
0.136197183098592, 0.126197183098592, 0.286197183098592), B = c(-0.294595070422536, 
-0.504595070422535, 0.125404929577464, 0.135404929577464, 0.275404929577465
), C = c(-0.277065727699531, -0.507065727699531, 0.282934272300469, 
0.0729342723004693, 0.122934272300469), D = c(-0.222699530516432, 
-0.132699530516432, -0.162699530516432, 0.127300469483568, -0.0126995305164321
), E = c(-0.246845657276995, -0.426845657276995, -0.186845657276995, 
0.133154342723005, 0.113154342723004)), row.names = c(NA, 5L), class = "data.frame")

Answer 1

Rfast::rownth gives the n th smallest value. 2nd highest value is 4th smallest value for a 5 column dataframe.

n <- 2
Rfast::rownth(as.matrix(df), rep(ncol(df) - n + 1, nrow(df)))
#[1] -0.2226995 -0.4268457  0.1361972  0.1331543  0.2754049

In base R with apply :

apply(df, 1, function(x) sort(x, decreasing = TRUE)[n])

Answer 2

Using collapse

library(collapse)
n <- 2
dapply(df, function(x) x[radixorder(x, decreasing = TRUE)][n], 
     MARGIN = 1) 
#[1] -0.2226995 -0.4268457  0.1361972  0.1331543  0.2754049

Or use fnth

-fnth(-t(df), n)
#       1          2          3          4          5 
#-0.2226995 -0.4268457  0.1361972  0.1331543  0.2754049