Add more arg types in Map transform (#305)

* Add more arg types in Map transform

* Fix show test

* Improve docstring

* Improve docstring

Author: juliohm

src/transforms/map.jl

@@ -2,19 +2,25 @@
 # Licensed under the MIT License. See LICENSE in the project root.
 # ------------------------------------------------------------------
 
-const TargetName = Union{Symbol,AbstractString}
-const PairWithTarget = Pair{<:Any,<:Pair{<:Function,<:TargetName}}
-const PairWithoutTarget = Pair{<:Any,<:Function}
-const MapPair = Union{PairWithTarget,PairWithoutTarget}
+# supported argument types
+const Callable = Union{Function,Type}
+const Target = Union{Symbol,AbstractString}
+const ColsCallableTarget = Pair{<:Any,<:Pair{<:Callable,<:Target}}
+const ColsCallable = Pair{<:Any,<:Callable}
+const CallableTarget = Pair{<:Callable,<:Target}
+const MapArg = Union{ColsCallableTarget,ColsCallable,CallableTarget,Callable}
 
 """
     Map(cols₁ => fun₁ => target₁, cols₂ => fun₂, ..., colsₙ => funₙ => targetₙ)
 
 Applies the `funᵢ` function to the columns selected by `colsᵢ` using 
 the `map` function and saves the result in a new column named `targetᵢ`.
+Types are also allowed in place of functions to construct objects with
+arguments from the columns.
 
 The column selection can be a single column identifier (index or name),
-a collection of identifiers or a regular expression (regex).
+a collection of identifiers or a regular expression (regex). It can also
+be ommited to apply the function to all columns.
 
 Passing a target column name is optional and when omitted a new name
 is generated by joining the function name with the selected column names.
@@ -28,11 +34,14 @@ Map([2, 3] => ((b, c) -> 2b + c))
 Map([:a, :c] => ((a, c) -> 2a * 3c) => :col1)
 Map(["c", "a"] => ((c, a) -> 3c / a) => :col1, "c" => tan)
 Map(r"[abc]" => ((a, b, c) -> a^2 - 2b + c) => "col1")
+Map(sin => "seno")
+Map(cos)
 ```
 
 ## Notes
 
-* Anonymous functions must be passed with parentheses as in the examples above;
+* Anonymous functions must be passed with parentheses as in the examples above
+
 * Some function names are treated in a special way, they are:
   * Anonymous functions: `#1` -> `f1`;
   * Composed functions: `outer ∘ inner` -> `outer_inner`;
@@ -41,20 +50,20 @@ Map(r"[abc]" => ((a, b, c) -> a^2 - 2b + c) => "col1")
 """
 struct Map <: StatelessFeatureTransform
   selectors::Vector{ColumnSelector}
-  funs::Vector{Function}
+  funs::Vector{Callable}
   targets::Vector{Union{Nothing,Symbol}}
 end
 
-Map() = throw(ArgumentError("cannot create Map transform without arguments"))
-
-function Map(pairs::MapPair...)
-  tuples = map(_extract, pairs)
-  selectors = [t[1] for t in tuples]
-  funs = [t[2] for t in tuples]
-  targets = [t[3] for t in tuples]
-  Map(selectors, funs, targets)
+function Map(args::MapArg...)
+  tups = map(_extract, args)
+  sels = [t[1] for t in tups]
+  funs = [t[2] for t in tups]
+  tars = [t[3] for t in tups]
+  Map(sels, funs, tars)
 end
 
+Map() = throw(ArgumentError("cannot create Map transform without arguments"))
+
 function applyfeat(transform::Map, feat, prep)
   cols = Tables.columns(feat)
   names = Tables.columnnames(cols)
@@ -76,13 +85,10 @@ function applyfeat(transform::Map, feat, prep)
   newfeat, nothing
 end
 
-_extract(p::PairWithTarget) = selector(first(p)), first(last(p)), Symbol(last(last(p)))
-_extract(p::PairWithoutTarget) = selector(first(p)), last(p), nothing
-
-_funname(fun::Base.Fix1) = "fix1_" * _funname(fun.f)
-_funname(fun::Base.Fix2) = "fix2_" * _funname(fun.f)
-_funname(fun::ComposedFunction) = _funname(fun.outer) * "_" * _funname(fun.inner)
-_funname(fun) = string(fun)
+_extract(arg::ColsCallableTarget) = selector(first(arg)), first(last(arg)), Symbol(last(last(arg)))
+_extract(arg::ColsCallable) = selector(first(arg)), last(arg), nothing
+_extract(arg::CallableTarget) = AllSelector(), first(arg), Symbol(last(arg))
+_extract(arg::Callable) = AllSelector(), arg, nothing
 
 function _makename(snames, fun)
   funname = _funname(fun)
@@ -91,3 +97,8 @@ function _makename(snames, fun)
   end
   Symbol(funname, :_, join(snames, "_"))
 end
+
+_funname(fun::Base.Fix1) = "fix1_" * _funname(fun.f)
+_funname(fun::Base.Fix2) = "fix2_" * _funname(fun.f)
+_funname(fun::ComposedFunction) = _funname(fun.outer) * "_" * _funname(fun.inner)
+_funname(fun) = string(fun)

test/shows.jl

@@ -279,14 +279,14 @@
     # compact mode
     iostr = sprint(show, T)
     @test iostr ==
-          "Map(selectors: ColumnSelector[:a, [:a, :b]], funs: Function[sin, $(nameof(fun))], targets: Union{Nothing, Symbol}[nothing, :c])"
+          "Map(selectors: ColumnSelector[:a, [:a, :b]], funs: Union{Function, Type}[sin, $(nameof(fun))], targets: Union{Nothing, Symbol}[nothing, :c])"
 
     # full mode
     iostr = sprint(show, MIME("text/plain"), T)
     @test iostr == """
     Map transform
     ├─ selectors: ColumnSelectors.ColumnSelector[:a, [:a, :b]]
-    ├─ funs: Function[sin, $(typeof(fun))()]
+    ├─ funs: Union{Function, Type}[sin, $(typeof(fun))()]
     └─ targets: Union{Nothing, Symbol}[nothing, :c]"""
   end
 

test/transforms/map.jl

@@ -96,6 +96,36 @@
   @test Tables.schema(n).names == (:fix2_hypot_a,)
   @test n.fix2_hypot_a == f.(t.a)
 
+  # function and target
+  f = (a, b, c, d) -> a + b + c + d
+  T = Map(f => "target")
+  n, c = apply(T, t)
+  @test Tables.schema(n).names == (:target,)
+  @test n.target == f.(t.a, t.b, t.c, t.d)
+
+  # function alone
+  f = (a, b, c, d) -> a + b + c + d
+  fname = replace(string(f), "#" => "f")
+  colname = Symbol(fname, :_a, :_b, :_c, :_d)
+  T = Map(f)
+  n, c = apply(T, t)
+  @test Tables.schema(n).names == (colname,)
+  @test Tables.getcolumn(n, colname) == f.(t.a, t.b, t.c, t.d)
+
+  # type and target
+  struct Foo a; b; c; d end
+  T = Map(Foo => "target")
+  n, c = apply(T, t)
+  @test Tables.schema(n).names == (:target,)
+  @test n.target == Foo.(t.a, t.b, t.c, t.d)
+
+  # type alone
+  struct Bar a; b; c; d end
+  T = Map(Bar)
+  n, c = apply(T, t)
+  @test Tables.schema(n).names == (:Bar_a_b_c_d,)
+  @test n.Bar_a_b_c_d == Bar.(t.a, t.b, t.c, t.d)
+
   # error: cannot create Map transform without arguments
   @test_throws ArgumentError Map()
 end