Remove first occurrence of elements in a vector from another vector

Question

I have a character vector, including some elements that are duplicates e.g.

v <- c("d09", "d11", "d13", "d01", "d02", "d10", "d13")

And another vector that includes single counts of those characters e.g.

x <- c("d10", "d11", "d13")

I want to remove only the first occurrence of each element in x from the 2nd vector v. In this example, d13 occurs in x and twice in v, but only the first match is removed from v and the duplicate is kept. Thus, I want to end up with:

"d09", "d01", "d02", "d13"

I've been trying various things e.g. z <- v[!(v %in% x)] but it keeps removing all instances of the characters in x, not just the first, so I end up with this instead:

"d09", "d01", "d02"

What can I do to only remove one instance of a duplicated element?

Answer 1

You can use match and negative indexing.

v[-match(x, v)]

produces

[1] "d09" "d01" "d02" "d13"

match only returns the location of the first match of a value, which we use to our advantage here.

Note that %in% and is.element are degenerate versions of match. Compare:

match(x, v)            # [1] 6 2 3
match(x, v) > 0        # [1] TRUE TRUE TRUE
x %in% v               # [1] TRUE TRUE TRUE
is.element(x, v)       # [1] TRUE TRUE TRUE

The last three are all the same, and are basically the coerced to logical version of the first (in fact, see code for %in% and is.element). In doing so you lose key information, which is the location of the first match of x in v and are left only knowing that x values exist in v.

The converse, v %in% x means something different from what you want, which is "which values in v are in x", which won't meet your requirement since all duplicate values will satisfy that condition.