any_is_na

NoteAdded in 0.0.8

What it does

Checks for usage of any(is.na(...)), NA %in% x, and NA %notin% x.

Why is this bad?

While both cases are valid R code, the base R function anyNA() is more efficient (both in speed and memory used).

Example

x <- c(1:10000, NA)
any(is.na(x))
NA %in% x
NA %notin% x

Use instead:

x <- c(1:10000, NA)
anyNA(x)
!anyNA(x)

References

See ?anyNA