Suggestion - Matrix Syntax for hcat (as well as vcat)

[Lincoln-Hannah]

function hvcat(rows::Tuple{Vararg{Int}}, d1::DataFrame, n1::NamedTuple, xs... ) 
    
    Rows = reshape([d1,n1,xs...], 2, Int64(length(xs)/2)+1) |> permutedims  |> eachrow

    vcat([hcat(d, DataFrame( 
        
                n[1] isa AbstractArray ? n : fieldnames(n)[1] => repeat([n[1]],nrow(d)) 
           ))

        for (d,n) in Rows ]...)

end
  

########    Examples     #################
[
    DataFrame(a=1:2,b=1:2) (;c="a")
    DataFrame(a=3:4,b=3:4) (;c="b")
] 

[
    DataFrame(a=1:2) (;b=1:2,c=11:12)
    DataFrame(a=3:4) (;b=3:4,c=13:14)
    DataFrame(a=5:7) (;b=5:7,c=15:17)
] 

[
    DataFrame(a=1:2,b=1:2)   DataFrame(c=1:2,d=1:2)
]

[bkamins]

Yes - we can consider adding hvcat in the next release.

[Lincoln-Hannah]

I often vcat a vector of DataFrames.
I don't think there is an easy way to add a column to show which vector element each dataframe came from.
This would be really useful. Something like:

vcat( [
    DataFrame(a=1:2,b=1:2) 
    DataFrame(a=3:4,b=3:4) 
]...,   
    extracolumn = :c => [  "df1",   "df2"  ]
)

output

:a   :b   :c
1    1    "df1"
2    2    "df1"
3    2    "df2"
4    2    "df2"

[bkamins]

julia> vcat(DataFrame(a=1:2,b=1:2), DataFrame(a=3:4,b=3:4),
           source = :c => [  "df1",   "df2"  ])
4×3 DataFrame
 Row │ a      b      c
     │ Int64  Int64  String
─────┼──────────────────────
   1 │     1      1  df1
   2 │     2      2  df1
   3 │     3      3  df2
   4 │     4      4  df2

[Lincoln-Hannah]

Thank you