Implement accessor methods for LabeledTable

Author: juliohm

src/StatsLearnModels.jl

@@ -34,6 +34,8 @@ include("learn.jl")
 export
   # labeled table
   LabeledTable,
+  predictors,
+  targets,
   label,
 
   # NearestNeighbors.jl

src/labeledtable.jl

@@ -23,6 +23,10 @@ function LabeledTable(table, names)
   LabeledTable{typeof(table)}(table, labs)
 end
 
+# -----------------
+# TABLES INTERFACE
+# -----------------
+
 Tables.istable(::Type{<:LabeledTable}) = true
 
 Tables.rowaccess(::Type{<:LabeledTable{T}}) where {T} = Tables.rowaccess(T)
@@ -35,6 +39,20 @@ Tables.columns(t::LabeledTable) = Tables.columns(t.table)
 
 Tables.columnnames(t::LabeledTable) = Tables.columnnames(t.table)
 
+# -----------------
+# ACCESSOR METHODS
+# -----------------
+
+Base.parent(t::LabeledTable) = t.table
+
+function predictors(t::LabeledTable)
+  cols = Tables.columns(t.table)
+  vars = Tables.columnnames(cols)
+  setdiff(vars, t.labels)
+end
+
+targets(t::LabeledTable) = t.labels
+
 # -----------
 # IO METHODS
 # -----------

test/runtests.jl

@@ -10,16 +10,34 @@ using Distributions: Binomial
 const SLM = StatsLearnModels
 
 @testset "StatsLearnModels.jl" begin
-  @testset "Basic" begin
-    # show method
-    x1 = rand(1:0.1:10, 100)
-    x2 = rand(1:0.1:10, 100)
-    y = 2x1 + x2
-    input = DataFrame(; x1, x2)
-    output = DataFrame(; y)
-    model = DecisionTreeClassifier()
-    fmodel = SLM.fit(model, input, output)
-    @test sprint(show, fmodel) == "FittedStatsLearnModel{DecisionTreeClassifier}"
+  @testset "LabeledTable" begin
+    # labels as symbols
+    t = (x1=rand(3), x2=rand(3), y=rand(Int, 3))
+    l = label(t, :y)
+    @test parent(l) == t
+    @test predictors(l) == [:x1, :x2]
+    @test targets(l) == [:y]
+
+    # labels as strings
+    t = (x1=rand(3), x2=rand(3), y=rand(Int, 3))
+    l = label(t, "y")
+    @test parent(l) == t
+    @test predictors(l) == [:x1, :x2]
+    @test targets(l) == [:y]
+
+    # multiple labels
+    t = (x1=rand(3), x2=rand(3), y1=rand(Int, 3), y2=rand(Int, 3))
+    l = label(t, ["y1", "y2"])
+    @test parent(l) == t
+    @test predictors(l) == [:x1, :x2]
+    @test targets(l) == [:y1, :y2]
+
+    # labels as regex
+    t = (x1=rand(3), x2=rand(3), y1=rand(Int, 3), y2=rand(Int, 3))
+    l = label(t, r"y")
+    @test parent(l) == t
+    @test predictors(l) == [:x1, :x2]
+    @test targets(l) == [:y1, :y2]
   end
 
   @testset "Models" begin
@@ -77,6 +95,16 @@ const SLM = StatsLearnModels
       accuracy = count(foutput.y .== output.y) / length(output.y)
       @test accuracy > 0
     end
+
+    # show method
+    x1 = rand(1:0.1:10, 100)
+    x2 = rand(1:0.1:10, 100)
+    y = 2x1 + x2
+    input = DataFrame(; x1, x2)
+    output = DataFrame(; y)
+    model = DecisionTreeClassifier()
+    fmodel = SLM.fit(model, input, output)
+    @test sprint(show, fmodel) == "FittedStatsLearnModel{DecisionTreeClassifier}"
   end
 
   @testset "Learn" begin