Merge pull request #67 from JuliaAI/featimp

Add feature importances support to iterated models

src/core.jl

@@ -155,3 +155,11 @@ MLJBase.transform(::EitherIteratedModel, fitresult, Xnew) =
 # here `fitresult` is a trained atomic machine:
 MLJBase.save(::EitherIteratedModel, fitresult) = MLJBase.serializable(fitresult)
 MLJBase.restore(::EitherIteratedModel, fitresult) = MLJBase.restore!(fitresult)
+
+# Feature importances
+function MLJBase.feature_importances(::EitherIteratedModel, fitresult, report)
+    # fitresult here is the curent state of the iterated machine 
+    # The line below will return `nothing` when the iteration model doesn't
+    # support feature_importances.
+    return MLJBase.feature_importances(fitresult)
+end

src/traits.jl

@@ -15,6 +15,7 @@ for trait in [:supports_weights,
               :is_pure_julia,
               :input_scitype,
               :output_scitype,
+              :reports_feature_importances,
               :target_scitype]
     quote
         # needed because traits are not always deducable from

test/core.jl

@@ -272,8 +272,10 @@ function MLJBase.fit(::EphemeralRegressor, verbosity, X, y)
     # if I serialize/deserialized `thing` then `id` below changes:
     id = objectid(thing)
     fitresult = (thing, id, mean(y))
-    return fitresult, nothing, NamedTuple()
+    report = (importances = [ftr => 1.0 for ftr in MLJBase.schema(X).names], )
+    return fitresult, nothing, report
 end
+
 function MLJBase.predict(::EphemeralRegressor, fitresult, X)
     thing, id, μ = fitresult
     return id == objectid(thing) ? fill(μ, nrows(X)) :
@@ -290,7 +292,12 @@ function MLJBase.restore(::EphemeralRegressor, serialized_fitresult)
     return (thing, id, μ)
 end
 
-@testset "save and restore" begin
+MLJBase.reports_feature_importances(::Type{<:EphemeralRegressor}) = true
+function MLJBase.feature_importances(::EphemeralRegressor, fitresult, report)
+    return report.importances
+end
+
+@testset "feature importances, save and restore" begin
     #https://github.com/JuliaAI/MLJ.jl/issues/1099
     X, y = (; x = rand(10)), fill(42.0, 3)
     controls = [Step(1), NumberLimit(2)]
@@ -302,12 +309,14 @@ end
     )
     mach = machine(imodel, X, y)
     fit!(mach, verbosity=0)
+    @test MLJBase.feature_importances(mach) == [:x => 1.0];
     io = IOBuffer()
     MLJBase.save(io, mach)
     seekstart(io)
     mach2 = machine(io)
     close(io)
     @test MLJBase.predict(mach2, (; x = rand(2))) ≈ fill(42.0, 2)
+
 end
 
 end

test/traits.jl

@@ -23,6 +23,7 @@ imodel = IteratedModel(model=model, measure=mae)
 @test output_scitype(imodel) == output_scitype(model)
 @test target_scitype(imodel) == target_scitype(model)
 @test constructor(imodel) == IteratedModel
+@test reports_feature_importances(imodel) == reports_feature_importances(model)
 
 end