Cover by Joseph Adler

Safari, the world’s most comprehensive technology and business learning platform.

Find the exact information you need to solve a problem on the fly, or go deeper to master the technologies and skills you need to succeed

Start Free Trial

No credit card required

O'Reilly logo

Sorting

Two final operations that you might find useful for analysis are sorting and ranking functions.

To sort the elements of an object, use the sort function:

> w <- c(5, 4, 7, 2, 7, 1)
> sort(w)
[1] 1 2 4 5 7 7

Add the decreasing=TRUE option to sort in reverse order:

> sort(w, decreasing=TRUE)
[1] 7 7 5 4 2 1

You can control the treatment of NA values by setting the na.last argument:

> length(w)
[1] 6
> length(w) <- 7
> # note that by default, NA.last=NA and NA values are not shown
> sort(w)
[1] 1 2 4 5 7 7
> # set NA.last=TRUE to put NA values last
> sort(w, na.last=TRUE)
[1]  1  2  4  5  7  7 NA
> # set NA.last=FALSE to put NA values first
> sort(w, na.last=FALSE)
[1] NA  1  2  4  5  7  7

Sorting data frames is somewhat nonintuitive. To sort a data frame, you need to create a permutation of the indices from the data frame and use these to fetch the rows of the data frame in the correct order. You can generate an appropriate permutation of the indices using the order function:

order(..., na.last = , decreasing = )

The order function takes a set of vectors as arguments. It sorts recursively by each vector, breaking ties by looking at successive vectors in the argument list. At the end, it returns a permutation of the indices of the vector corresponding to the sorted order. (The arguments na.last and decreasing work the same way as they do for sort.) To see what this means, let’s use a simple example. First, we’ll define a vector with two elements out of order:

> v <- c(11, 12, 13, 15, 14)

You can ...

Find the exact information you need to solve a problem on the fly, or go deeper to master the technologies and skills you need to succeed

Start Free Trial

No credit card required