if_else missing argument does not work when using %in% to evaluate.

[Raesu]

This function could refer to the missing argument to fill in missing values. It works except for when %in% is used in the condition to check.

---

x <- c(NA, 1,2,3)

# Expected behavior, NA passes through
if_else(x == 2, TRUE, FALSE)

# NA is converted to FALSE
if_else(x %in% c(2), TRUE, FALSE)

# NA is converted to FALSE, `missing` argument has no impact
if_else(x %in% c(2), TRUE, FALSE, missing=NA)

[DavisVaughan]

The conversion of the NA to FALSE happens with %in%, and is unrelated to dplyr code

x <- c(NA, 1,2,3)

x %in% 2
#> [1] FALSE FALSE  TRUE FALSE

So there is nothing for dplyr to do here. Converting NA to FALSE is just a known side effect of how %in% and match() work.

[DavisVaughan]

If you really want NAs to pass through, you can use vctrs::vec_in() with na_equal = FALSE

vctrs::vec_in(c(NA, 1, 2, 3), 2, na_equal = FALSE)
#> [1]    NA FALSE  TRUE FALSE

But for the most part that is considered "advanced usage"

[Raesu]

Helpful, I did not know that. Thanks.