The Art of R Programming

We can sort by word frequency in a similar manner.

1 # orders the output of findwords() by word frequency
2 freqwl <- function(wrdlst) {
3 freqs <- sapply(wrdlst,length) # get word frequencies
4 return(wrdlst[order(freqs)])
5 }

In line 3, we are using the fact that each element inwrdlstis a vector

of numbers representing the positions in our input file at which the given

word is found. Callinglength()on that vector gives us the number of times

the given word appears in that file. The result of callingsapply()will be the

vector of word frequencies.

We could usesort()here again, butorder()is more direct. This latter

function returns the indices of a sorted vector with respect to the original

vector. Here’s an example:

> x <- c(12,5,13,8)

> order(x)

[1]2413

The output here indicates thatx[2]is the smallest element inx,x[4]

is the second smallest, and so on. In our case, we useorder()to determine

which word is least frequent, second least frequent, and so on. Plugging

these indices into our word list gives us the same word list but in order of

frequency.

Let’s check it.

> freqwl(wl)

$here

[1] 2

$means

[1] 3

$first

[1] 6

...

$that

[1] 4 40

$`in`

[1] 8 15

$line

[1] 10 24

...

$this

Lists 97