Sort two columns, conditionally select values, then run cumsum frequency

Question

My data looks like this

a   b   c
1   1   0
1   2   8
2   1   0
2   2   2
3   1   3
3   2   3
4   1   7
4   2   4
5   1   3
5   2   5
6   1   1
6   2   8
7   1   1
7   2   2

I want to sort columns a and c so that every even-numbered row in column c is the largest number for every pair in column a. Then I want to take these values and store them in a new object. It should look something like this.

a   c   b
1   8   2
2   2   2
3   3   2
4   7   1
5   5   2
6   8   2
7   2   2

Answer 1

With data.table package you can sort your data by reference using setorder or setkey (without the need of creating copies using <- function)

library(data.table)
setorder(setDT(df), a, c)[]
#     a b c
#  1: 1 1 0
#  2: 1 2 8
#  3: 2 1 0
#  4: 2 2 2
#  5: 3 1 3
#  6: 3 2 3
#  7: 4 1 7
#  8: 4 2 4
#  9: 5 1 3
# 10: 5 2 5
# 11: 6 1 1
# 12: 6 2 8
# 13: 7 1 1
# 14: 7 2 2

Then you can achieve your goal in various simple ways, for example

df[duplicated(a)]
#    a b c
# 1: 1 2 8
# 2: 2 2 2
# 3: 3 2 3
# 4: 4 2 4
# 5: 5 2 5
# 6: 6 2 8
# 7: 7 2 2

Or maybe

df[, tail(.SD, 1), a]
#    a b c
# 1: 1 2 8
# 2: 2 2 2
# 3: 3 2 3
# 4: 4 2 4
# 5: 5 2 5
# 6: 6 2 8
# 7: 7 2 2

Or

df[, .SD[2], a]
#    a b c
# 1: 1 2 8
# 2: 2 2 2
# 3: 3 2 3
# 4: 4 2 4
# 5: 5 2 5
# 6: 6 2 8
# 7: 7 2 2

P.S. If you want to change the order of the columns, you can also do this by reference using setcolorder function, e.g.,

setcolorder(df, c("a", "c", "b"))