Find the missing values in Julia like R's is.n

Question

The Julia 1.0.0 documentation says this about missing values in Julia and R:

In Julia, missing values are represented by the missing object rather than by NA. Use ismissing(x) instead of isna(x). The skipmissing function is generally used instead of na.rm=TRUE (though in some particular cases functions take a skipmissing argument).

Here is example code in R that I would like to duplicate in Julia:

> v = c(1, 2, NA, 4)
> is.na(v)
[1] FALSE FALSE  TRUE FALSE

(First note that is.na is the R function's correct spelling, not isna as shown in the quote above, but that is not my point.)

If I follow the documentation's suggestion to use ismissing in Julia, I get a different type of result than in R.

julia> v = [1, 2, missing, 4]
4-element Array{Union{Missing, Int64},1}:
 1
 2
  missing
 4

# Note that based on R, I was expecting: `false false true false` 
# though obviously in a different output format.
julia> ismissing(v)
false

To duplicate the R code, I seem to have to do something like:

julia> [ismissing(x) for x in v]
4-element Array{Bool,1}:
 false
 false
  true
 false

That works, but it is not as succinct as is.na in R. Maybe I am missing something.

I also tried:

julia> ismissing(v[:])
false

julia> ismissing(v[1:end])
false

Any suggestions?

Answer 1

You can broadcast ismissing with .:

julia> v = [1, 2, missing, 4]
4-element Array{Union{Missing, Int64},1}:
 1
 2
  missing
 4

julia> ismissing.(v)
4-element BitArray{1}:
 false
 false
  true
 false