I agree with the above, although it seems to me like you are just trying to vertically concatenate the DataFrames rather than join
them?
julia> vcat(test1, test2, cols = :union)
6×3 DataFrame
Row │ ID Name Argh
│ Int64 String Float64?
─────┼──────────────────────────
1 │ 1 A 5.0
2 │ 2 B 62.1
3 │ 3 C 89.0
4 │ 4 D missing
5 │ 5 E missing
6 │ 6 F missing
(note the cols = :union
argument which fills in the missing Argh
column in test2
with missing
)