Creates {prefix}_any (1 if any are 1), {prefix}_count (number of 1s),
and {prefix}_all (1 if all are 1) from a set of binary columns. All three
are NA when every source column is NA.
Examples
d <- data.frame(x = c(1, 0, NA), y = c(1, 1, NA), z = c(0, 1, NA))
count_binary(d, "xyz", x, y, z)
#> x y z xyz_any xyz_count xyz_all
#> 1 1 1 0 1 2 0
#> 2 0 1 1 1 2 0
#> 3 NA NA NA NA NA NA