Dict from dataframe

General Usage
,
1

df = DataFrame(ID=1:4,X=11:14, Y=10:13)

indexing to one colum worked using this - Dict(pairs(df.x))
I wanted to convert this to Dict in julia to get like

1 => [ 11,10]
2 => [12,11]
3 => [13,12]
4 => [14,13]

Tried Dict(pairs([df.x,df.y])), shows error

Is there any other way to get this done

2 
julia> Dict(df.ID[i] => [df.X[i], df.Y[i]] for i in 1:nrow(df))
Dict{Int64, Vector{Int64}} with 4 entries:
  4 => [14, 13]
  2 => [12, 11]
  3 => [13, 12]
  1 => [11, 10]
2 Likes
3

Another alternative:

julia> Dict(i => collect(u) for (i,u) in enumerate(zip(df.X,df.Y)))
Dict{Int64, Vector{Int64}} with 4 entries:
  4 => [14, 13]
  2 => [12, 11]
  3 => [13, 12]
  1 => [11, 10]
3 Likes
4

And using Julia broadcasting for compactness:

 Dict(df.ID .=> eachrow([df.X df.Y]))

  4 => [14, 13]
  2 => [12, 11]
  3 => [13, 12]
  1 => [11, 10]
6 Likes
5

This might also be useful to you. It will work even if the number of columns change. A DataFrameRow works just like a Vector except you can also index into it by column name.

julia> d = Dict(df.ID .=> eachrow(df[:, Not(:ID)]))
Dict{Int64, DataFrameRow{DataFrame, DataFrames.Index}} with 4 entries:
  4 => DataFrameRow…
  2 => DataFrameRow…
  3 => DataFrameRow…
  1 => DataFrameRow…

julia> d[1]
DataFrameRow
 Row │ X      Y
     │ Int64  Int64
─────┼──────────────
   1 │    11     10

julia> d[1][2]
10

julia> d[1].Y
10
2 Likes
6 
julia> Dict(zip(df.ID,zip(df.X, df.Y)))
Dict{Int64, Tuple{Int64, Int64}} with 4 entries:
  4 => (14, 13)
  2 => (12, 11)
  3 => (13, 12)
  1 => (11, 10)

julia> Dict(zip(df.ID,zip(eachcol(df[:,Not(:ID)])...)))
Dict{Int64, Tuple{Int64, Int64}} with 4 entries:
  4 => (14, 13)
  2 => (12, 11)
  3 => (13, 12)
  1 => (11, 10)
1 Like