How to order data horizontally (column-wise)?

 

How to order data horizontally (column-wise)?

I have data in the following manner. For each category x taking values a, b, or c, y is an integer.

x1	y1	x2	y2	x3	y3	x4	y4
b	7	b	4	c	4	b	1
a	5	a	9	b	1	b	7

I'd like to sort the data in ascending order column-wise (horizontally) when x = "b". Desired output:

x1	y1	x2	y2	x3	y3	x4	y4
b	1	b	4	b	7	c	4
b	1	b	7	a	5	a	9

I tried using a bubble sort algorithm to arrange the y values, but unable to conditionally do it when x = "b". Any help is appreciated!

Answer

A little awkward, but works.

lapply(asplit(dat, 1), matrix, ncol=2, byrow=TRUE) |>
  lapply(as.data.frame) |>
  sapply(\(x) {
    s <- split(x, x[, 1] == 'b')
    s$`TRUE` <- s$`TRUE`[order(s$`TRUE`[, 2]), ]
    r <- rbind(s$`TRUE`, s$`FALSE`)
    asplit(r, 1) |> unlist()
  }) |> t() |> as.data.frame() |> setNames(names(dat))
#   x1 y1 x2 y2 x3 y3 x4 y4
# 1  b  1  b  4  b  7  c  4
# 2  b  1  b  7  a  5  a  9

---

Data:

dat <- yaj::read.table2('x1 y1  x2  y2  x3  y3  x4  y4
b   7   b   4   c   4   b   1
a   5   a   9   b   1   b   7')