Support joining dataframes on columns with different left/right names #1312
Conversation
src/abstractdataframe/join.jl
Outdated
| @@ -6,21 +6,34 @@ | |||
| similar_missing(dv::AbstractArray{T}, dims::Union{Int, Tuple{Vararg{Int}}}) where {T} = | |||
| fill!(similar(dv, Union{T, Missing}, dims), missing) | |||
|
|
|||
| # OnType | |||
| const OT = Union{Symbol, NTuple{2,Symbol}, Pair{Symbol,Symbol}} | |||
There was a problem hiding this comment.
Better spell it in full: OnType.
src/abstractdataframe/join.jl
Outdated
|
|
||
| ### Arguments | ||
|
|
||
| * `df1`, `df2` : the two AbstractDataFrames to be joined | ||
|
|
||
| ### Keyword Arguments | ||
|
|
||
| * `on` : a Symbol or Vector{Symbol}, the column(s) used as keys when | ||
| joining; required argument except for `kind = :cross` | ||
| * `on` : A column, or vector of columns to join df1 and df2 on; If the column(s) |
There was a problem hiding this comment.
Use period rather than semicolon. Also add backticks around (left, right) and (left => right), and "Tuple" should either be in backticks and in singular, or be in lowercase.
Could you also add an example?
test/join.jl
Outdated
| DataFrame(id = 3:7, sid = string.(3:7), SID = string.(3:7)) | ||
| @test join(left, right, on = [(:id, :ID), (:sid, :SID)]) == | ||
| DataFrame(id = 3:7, sid = string.(3:7)) | ||
| @test join(left, right, on = [(:id => :ID), (:sid => :SID)]) == |
There was a problem hiding this comment.
Parentheses aren't needed, right?
| @@ -371,4 +371,21 @@ module TestJoin | |||
| @test levels(join(A, B, on=:b, kind = :semi)[:b]) == ["d", "c", "b", "a"] | |||
| @test levels(join(B, A, on=:b, kind = :semi)[:b]) == ["a", "b", "c"] | |||
| end | |||
|
|
|||
| @testset "join on columns with different left/right names" begin | |||
There was a problem hiding this comment.
Can you also tests different kinds of joins?
There was a problem hiding this comment.
Great idea. Your suspicion
Looks mostly good, but I wonder how it can work given that left_on isn't used anywhere...
was very astute : )
src/abstractdataframe/join.jl
Outdated
| on::Union{Symbol, Vector{Symbol}} = Symbol[], | ||
| kind::Symbol = :inner) | ||
| ``` | ||
| Join two DataFrames |
There was a problem hiding this comment.
data frames or `DataFrame` objects
| end | ||
| end | ||
|
|
||
| DataFrameJoiner(dfl::DF1, dfr::DF2, on::Union{Symbol,Vector{Symbol}}) where {DF1<:AbstractDataFrame, DF2<:AbstractDataFrame} = | ||
| DataFrameJoiner(dfl::DF1, dfr::DF2, on::Union{<:OnType, AbstractVector{<:OnType}}) where | ||
| {DF1<:AbstractDataFrame, DF2<:AbstractDataFrame} = |
There was a problem hiding this comment.
I guess that's the indentation convention for this kind of situation? I'm not used to it yet.
There was a problem hiding this comment.
I'm not sure if there is a convention, but this only works if the line break is after the where clause. If you line break before the where then it results in an error
There was a problem hiding this comment.
Yeah, that's annoying. But I was wondering whether we should use four spaces, or align with arguments. Hard to decide.
Banishes #1297 to Closed