How to order column of data.frame by another variable

Question

mydata <- data.frame(id = c(rep(1, 3), rep(2, 3), rep(3, 3)), 
                      score = c(c(1, 2, 3), c(3, 2, 1), c(1, 3, 2)),
                      location = c(rep(c("X", "Y", "Z"), 3)))
> mydata
  id score location
1  1     1        X
2  1     2        Y
3  1     3        Z
4  2     3        X
5  2     2        Y
6  2     1        Z
7  3     1        X
8  3     3        Y
9  3     2        Z

I would like to sort my data.frame according to score from smallest to largest for each id.

Simplying ordering by score ignores the id column.

> mydata[with(mydata, order(score)),]
  id score location
1  1     1        X
6  2     1        Z
7  3     1        X
2  1     2        Y
5  2     2        Y
9  3     2        Z
3  1     3        Z
4  2     3        X
8  3     3        Y

Essentially, I want my output to be

  id score location
1  1     1        X
2  1     2        Y
3  1     3        Z
4  2     1        Z
5  2     2        Y
6  2     3        X
7  3     1        X
8  3     2        Z
9  3     3        Y

score 3 · Answer 1 · answered Aug 15 '17 at 20:45

You can use dplyr package:

library(dplyr)
mydata %>% arrange(id,score)

#   id score location 
# 1  1     1        X 
# 2  1     2        Y 
# 3  1     3        Z 
# 4  2     1        Z 
# 5  2     2        Y 
# 6  2     3        X 
# 7  3     1        X 
# 8  3     2        Z 
# 9  3     3        Y

score 3 · Accepted Answer · answered Aug 15 '17 at 20:57

Using base R only.

mydata[order(mydata$id, mydata$score), ]
  id score location
1  1     1        X
2  1     2        Y
3  1     3        Z
6  2     1        Z
5  2     2        Y
4  2     3        X
7  3     1        X
9  3     2        Z
8  3     3        Y

How to order column of data.frame by another variable

2 Answers2