Add Coerce

src/TableTransforms.jl

@@ -36,6 +36,7 @@ export
   DropMissing,
   Rename,
   Identity,
+  Coerce,
   Center,
   Scale,
   MinMax,

src/transforms.jl

@@ -216,6 +216,7 @@ include("transforms/select.jl")
 include("transforms/filter.jl")
 include("transforms/rename.jl")
 include("transforms/identity.jl")
+include("transforms/coerce.jl")
 include("transforms/center.jl")
 include("transforms/scale.jl")
 include("transforms/zscore.jl")

src/transforms/coerce.jl

@@ -0,0 +1,34 @@
+# ------------------------------------------------------------------
+# Licensed under the MIT License. See LICENSE in the project root.
+# ------------------------------------------------------------------
+
+"""
+    Coerce(pairs, tight=false, verbosity=1)
+
+Return a copy of the table, ensuring that the element scitypes of the columns match the new specification.
+
+Valid specifications (more to be added):
+
+(1) one or more column_name=>Scitype pairs
+"""
+struct Coerce{P, T, V} <: Transform
+  pairs::P
+  tight::T
+  verbosity::V
+end
+
+Coerce(pair...; tight=false, verbosity=1) = Coerce(pair, tight, verbosity)
+
+isrevertible(::Type{<:Coerce}) = true
+
+function apply(transform::Coerce, table)
+  newtable = ScientificTypes.coerce(table, transform.pairs...; tight=transform.tight, verbosity=transform.verbosity)
+
+  scitypes = [ScientificTypes.elscitype(x) for x in Tables.columns(table)]
+  colnames = Tables.columnnames(table)
+  pairs = [Pair(i, j) for (i, j) in zip(colnames, scitypes)]
+
+  return newtable, pairs
+end
+
+revert(transform::Coerce, newtable, cache) = ScientificTypes.coerce(newtable, cache...)

test/Project.toml

@@ -6,6 +6,7 @@ LinearAlgebra = "37e2e46d-f89d-539d-b4ee-838fcccc9c8e"
 Plots = "91a5bcdd-55d7-5caf-9e0b-520d859cae80"
 Random = "9a3f8284-a2c9-5f02-9a11-845980a1fd5c"
 ReferenceTests = "324d217c-45ce-50fc-942e-d289b448e8cf"
+ScientificTypes = "321657f4-b219-11e9-178b-2701a2544e81"
 Statistics = "10745b16-79ce-11e8-11f9-7d13ad32a3b2"
 Tables = "bd369af6-aec1-5ad0-b16a-f7cc5008161c"
 Test = "8dfed614-e22c-5e08-85e1-65c5234f0b40"

test/runtests.jl

@@ -6,6 +6,9 @@ using LinearAlgebra
 using Statistics
 using Test, Random, Plots
 using ReferenceTests, ImageIO
+# WARNING: replacing the following line with `using ScientificTypes` will
+# cause a conflicting import and break the tests
+import ScientificTypes: Count, elscitype
 
 # set default configurations for plots
 gr(ms=2, mc=:black, aspectratio=:equal,

test/transforms.jl

@@ -685,6 +685,22 @@
     @test t == tₒ
   end
 
+  @testset "Coerce" begin
+    x1 = [1.0, 2.0, 3.0, 4.0, 5.0]
+    x2 = [1.0, 2.0, 3.0, 4.0, 5.0]
+    x3 = [5.0, 5.0, 5.0, 5.0, 5.0]
+    t = TypedTables.Table(;x1, x2, x3)
+
+    # apply test
+    n, c = apply(Coerce(:x1=>Count, :x2=>Count), t)
+    @test elscitype(n.x1) == Count && elscitype(n.x2) == Count
+
+    # revert test
+    n, c = apply(Coerce(:x1=>Count, :x2=>Count), t)
+    tₒ = revert(Coerce(:x1=>Count, :x2=>Count), n, c)
+    @test t == tₒ
+  end
+
   @testset "Center" begin
     # using rng for reproducible results
     x = rand(rng, Normal(2, 1), 4000)