Count unique combinations

n_distinct() counts the number of unique/distinct combinations in a set of one or more vectors. It's a faster and more concise equivalent to nrow(unique(data.frame(...))).

Usage

n_distinct(..., na.rm = FALSE)

Arguments

...

Unnamed vectors. If multiple vectors are supplied, then they should have the same length.

na.rm

If TRUE, exclude missing observations from the count. If there are multiple vectors in ..., an observation will be excluded if any of the values are missing.

Value

A single number.

Examples

x <- c(1, 1, 2, 2, 2)
n_distinct(x)
#> [1] 2

y <- c(3, 3, NA, 3, 3)
n_distinct(y)
#> [1] 2
n_distinct(y, na.rm = TRUE)
#> [1] 1

# Pairs (1, 3), (2, 3), and (2, NA) are distinct
n_distinct(x, y)
#> [1] 3

# (2, NA) is dropped, leaving 2 distinct combinations
n_distinct(x, y, na.rm = TRUE)
#> [1] 2

# Also works with data frames
n_distinct(data.frame(x, y))
#> [1] 3