CODEWITHSUNDEEP
r
Bhail
Bhail

Reputation: 407

Is sort( ) poor at rearraing because it returns the order as numeric vector?

I quickly learned that arrange() from {dplyr} is a better for my analysis than sort(). Just curious it is because the object returned by sort() function is of numeric class. Example: 

df <- data.frame(a=round(rpois(10,10)), b=LETTERS[1:10], stringsAsFactors = F)
df
    a b
1  13 A
2   7 B
3  11 C
4   6 D
5  10 E
6  10 F
7  10 G
8   6 H
9  11 I
10  9 J

I found out that arrange() is good at rearranging if I want to sort the data frame.

arrange(df, desc(a))
    a b
1  13 A
2  11 C
3  11 I
4  10 E
5  10 F
6  10 G
7   9 J
8   7 B
9   6 D
10  6 H

If I try to order the original data frame, what is sort doing?

df[sort(df[,1]),]
      a    b
6    10    F
6.1  10    F
7    10    G
9    11    I
10    9    J
10.1  9    J
10.2  9    J
NA   NA <NA>
NA.1 NA <NA>
NA.2 NA <NA>

So I am think it is generally not a good idea to rearrange with a numeric vector.

sort(df[,1])
[1]  6  6  7  9 10 10 10 11 11 13
class(sort(df[,1]))
[1] "numeric"

I would like to know what happened when I tried to rearrange with sort(). Thanks.

Upvotes: 1

Views: 53

Answers (1)

akrun
akrun

Reputation: 887118

We can use order to provide the index.

df[order(-df$a),]

If we are using sort, use the index.return option as TRUE

df[sort(df$a, decreasing = TRUE, index.return = TRUE)$ix,]

Upvotes: 2

Related Questions