86

What are good examples of when seq_along will work, but seq will produce unintended results?

From the documentation of ?seq we have:

Note that it dispatches on the class of the first argument irrespective of argument names. This can have unintended consequences if it is called with just one argument intending this to be taken as along.with: it is much better to use seq_along in that case.

Henrik
  • 65,555
  • 14
  • 143
  • 159
Ricardo Saporta
  • 54,400
  • 17
  • 144
  • 178

2 Answers2

137

This should make the difference clear. Basically, seq() acts like seq_along() except when passed a vector of length 1, in which case it acts like seq_len(). If this ever once bites you, you'll never use seq() again!

a <- c(8, 9, 10)
b <- c(9, 10)
c <- 10

seq_along(a)
# [1] 1 2 3
seq_along(b)
# [1] 1 2
seq_along(c)
# [1] 1

seq(a)
# [1] 1 2 3
seq(b)
# [1] 1 2
seq(c)
# [1]  1  2  3  4  5  6  7  8  9 10

It's probably worth noting that sample() exhibits similarly crummy behavior:

sample(a)
# [1] 10  8  9
sample(b)
# [1]  9 10
sample(c)
# [1]  8  7  9  3  4  1  6 10  2  5
Josh O'Brien
  • 159,210
  • 26
  • 366
  • 455
26

If the input to seq is length 1 then the outputs between seq and seq_along will be different

x <- 5
for(i in seq(x)){
    print(x[i])
}
#[1] 5
#[1] NA
#[1] NA
#[1] NA
#[1] NA

for(i in seq_along(x)){
    print(x[i])
}
#[1] 5

We also see a difference if the input is a vector of Dates

x <- Sys.Date() + 1:5
seq(x)
#Error in seq.Date(x) : 'from' must be of length 1
seq_along(x)
#[1] 1 2 3 4 5
Dason
  • 60,663
  • 9
  • 131
  • 148