CODEWITHSUNDEEP
dataframejulia
AVA
AVA

Reputation: 2558

How to rename column name in DataFrames (version 0.19) while combining?

DataFrame:

julia> df4
135×4 DataFrame
│ Row │ County      │ Year  │ Female │ Male   │
│     │ String      │ Int64 │ Int64⍰ │ Int64⍰ │
├─────┼─────────────┼───────┼────────┼────────┤
│ 1   │ Asotin      │ 2008  │ 1      │ 0      │
│ 2   │ Asotin      │ 2009  │ 0      │ 0      │
│ 3   │ Asotin      │ 2010  │ 0      │ 0      │
│ 4   │ Asotin      │ 2011  │ 0      │ 0      │
│ 5   │ Asotin      │ 2012  │ 0      │ 0      │
│ 6   │ Benton      │ 2008  │ 1      │ 0      │
│ 7   │ Benton      │ 2009  │ 0      │ 0      │
│ 8   │ Benton      │ 2010  │ 0      │ 0      │
│ 9   │ Benton      │ 2011  │ 0      │ 0      │
│ 10  │ Benton      │ 2012  │ 0      │ 0      │
│ 11  │ Chelan      │ 2008  │ 1      │ 0      │
│ 12  │ Chelan      │ 2009  │ 1      │ 0      │
│ 13  │ Chelan      │ 2010  │ 0      │ 1      │
│ 14  │ Chelan      │ 2011  │ 0      │ 0      │
│ 15  │ Chelan      │ 2012  │ 0      │ 2      │
│ 16  │ Clallam     │ 2008  │ 0      │ 0      │
│ 17  │ Clallam     │ 2009  │ 0      │ 0      │
│ 18  │ Clallam     │ 2010  │ 0      │ 0      │
│ 19  │ Clallam     │ 2011  │ 1      │ 1      │
│ 20  │ Clallam     │ 2012  │ 0      │ 0      │
│ 21  │ Clark       │ 2008  │ 0      │ 1      │
⋮
│ 114 │ Thurston    │ 2011  │ 0      │ 0      │
│ 115 │ Thurston    │ 2012  │ 0      │ 0      │
│ 116 │ Walla Walla │ 2008  │ 0      │ 0      │
│ 117 │ Walla Walla │ 2009  │ 0      │ 1      │
│ 118 │ Walla Walla │ 2010  │ 0      │ 0      │
│ 119 │ Walla Walla │ 2011  │ 0      │ 0      │
│ 120 │ Walla Walla │ 2012  │ 0      │ 0      │
│ 121 │ Whatcom     │ 2008  │ 0      │ 0      │
│ 122 │ Whatcom     │ 2009  │ 1      │ 0      │
│ 123 │ Whatcom     │ 2010  │ 0      │ 1      │
│ 124 │ Whatcom     │ 2011  │ 1      │ 1      │
│ 125 │ Whatcom     │ 2012  │ 0      │ 1      │
│ 126 │ Whitman     │ 2008  │ 0      │ 0      │
│ 127 │ Whitman     │ 2009  │ 0      │ 0      │
│ 128 │ Whitman     │ 2010  │ 0      │ 1      │
│ 129 │ Whitman     │ 2011  │ 0      │ 0      │
│ 130 │ Whitman     │ 2012  │ 0      │ 0      │
│ 131 │ Yakima      │ 2008  │ 0      │ 0      │
│ 132 │ Yakima      │ 2009  │ 0      │ 1      │
│ 133 │ Yakima      │ 2010  │ 1      │ 2      │
│ 134 │ Yakima      │ 2011  │ 0      │ 3      │
│ 135 │ Yakima      │ 2012  │ 0      │ 1      │

Without column renaming, groupby and combine, as follows:

julia> let 
           data = @pipe df4 |>
               groupby(_, :Year) |>
               combine(_, :Female => sum, :Male => sum)
       end
5×3 DataFrame
│ Row │ Year  │ Female_sum │ Male_sum │
│     │ Int64 │ Int64      │ Int64    │
├─────┼───────┼────────────┼──────────┤
│ 1   │ 2008  │ 7          │ 13       │
│ 2   │ 2009  │ 8          │ 11       │
│ 3   │ 2010  │ 4          │ 21       │
│ 4   │ 2011  │ 7          │ 22       │
│ 5   │ 2012  │ 5          │ 31       │

Renaming while combining throws error, as follows:

julia> let 
           data = @pipe df4 |>
               groupby(_, :Year) |>
               combine(_, :Female => sum => :Female, :Male => sum => :Male)
       end
ERROR: MethodError: objects of type Pair{typeof(sum),Symbol} are not callable
Stacktrace:
 [1] fun at /opt/julia/julia-1.4.1/share/julia/stdlib/v1.4/packages/DataFrames/yH0f6/src/groupeddataframe/grouping.jl:671 [inlined]
 [2] do_call(::DataFrames.var"#fun#173"{Pair{typeof(sum),Symbol}}, ::GroupedDataFrame{DataFrame}, ::Array{Union{Missing, Int64},1}, ::Int64) at /opt/julia/julia-1.4.1/share/julia/stdlib/v1.4/packages/DataFrames/yH0f6/src/groupeddataframe/grouping.jl:465
 [3] (::DataFrames.var"#174#178"{GroupedDataFrame{DataFrame}})(::Pair{Symbol,Pair{typeof(sum),Symbol}}) at /opt/julia/julia-1.4.1/share/julia/stdlib/v1.4/packages/DataFrames/yH0f6/src/groupeddataframe/grouping.jl:698
 [4] map(::DataFrames.var"#174#178"{GroupedDataFrame{DataFrame}}, ::Tuple{Pair{Symbol,Pair{typeof(sum),Symbol}},Pair{Symbol,Pair{typeof(sum),Symbol}}}) at ./tuple.jl:158
 [5] _combine(::Tuple{Pair{Symbol,Pair{typeof(sum),Symbol}},Pair{Symbol,Pair{typeof(sum),Symbol}}}, ::GroupedDataFrame{DataFrame}) at /opt/julia/julia-1.4.1/share/julia/stdlib/v1.4/packages/DataFrames/yH0f6/src/groupeddataframe/grouping.jl:683
 [6] combine(::Tuple{Pair{Symbol,Pair{typeof(sum),Symbol}},Pair{Symbol,Pair{typeof(sum),Symbol}}}, ::GroupedDataFrame{DataFrame}) at /opt/julia/julia-1.4.1/share/julia/stdlib/v1.4/packages/DataFrames/yH0f6/src/groupeddataframe/grouping.jl:437
 [7] combine(::GroupedDataFrame{DataFrame}, ::Pair{Symbol,Pair{typeof(sum),Symbol}}, ::Pair{Symbol,Pair{typeof(sum),Symbol}}) at /opt/julia/julia-1.4.1/share/julia/stdlib/v1.4/packages/DataFrames/yH0f6/src/groupeddataframe/grouping.jl:444
 [8] top-level scope at REPL[48]:2

Please guide me in renaming the column name while combining in DataFrame0.19.

Upvotes: 1

Views: 88

Answers (1)

Nils Gudat
Nils Gudat

Reputation: 13800

As I said on one of your previous questions here, you really shouldn't be using DataFrames 0.19 at this point, the current release is 1.2 and a lot of things have changed since then.

One of these changes was to change the syntax for combine and transform to a uniform pattern of :sourcecol => transformation => :targetcol. In previous versions, the syntax was targetcol = :sourcecol => transformation):

(@v1.6) pkg> activate --temp
  Activating new environment at `/tmp/jl_59ViXd/Project.toml`

(jl_59ViXd) pkg> add [email protected]

julia> using DataFrames

julia> df = DataFrame(a = rand('a':'b', 10), b = rand(10))

julia> combine(groupby(df, :a), newcol = :b => sum)
2×2 typename(DataFrame)
│ Row │ a    │ newcol  │
│     │ Char │ Float64 │
├─────┼──────┼─────────┤
│ 1   │ 'a'  │ 1.94586 │
│ 2   │ 'b'  │ 3.00895 │

But let me reiterate that it's not a good idea to learn outdated and currently unsupported DataFrames syntax if you want to continue using the package!

Upvotes: 3

Related Questions