ArgumentError: column(s) c are missing from argument(s) 1, and column(s) a are missing from argument(s) 2

Question

I would like to row bind two dataframes with different columns like bind_rows in R. I tried the same with vcat as in this question: Julia equivalent of dplyr's bind_cols and bind_rows, but I get an error. Here some reproducible code:

using DataFrames

df1 = DataFrame(a = 1, b = 1)
df2 = DataFrame(b = 1, c = 1)
vcat(df1, df2)

Output:

ArgumentError: column(s) c are missing from argument(s) 1, and column(s) a are missing from argument(s) 2

I don't know why this error happens, because in the answer of the question it does work. So I was wondering if anyone knows why this error happens and if this is the best way to bind two dataframes in Julia?

Answer 1

The error occurs because those DataFrames have columns with different names (technically, properties). Use the cols=:union keyword argument:

vcat(df1, df2, cols=:union)
# 2×3 DataFrame
#  Row │ a        b      c
#      │ Int64?   Int64  Int64?
# ─────┼─────────────────────────
#    1 │       1      1  missing
#    2 │ missing      1        1

union holds a union of columns. cols have more options. check them out here if you are interested.