Improve test coverage

src/utilities.jl

@@ -135,20 +135,6 @@ function recursive_setproperty!(obj, ex::Expr, value)
     return recursive_setproperty!(last_obj, field, value)
 end
 
-"""
-    check_dimensions(X, Y)
-
-Internal function to check two arrays have the same shape.
-
-"""
-@inline function check_dimensions(X, Y)
-    size(X)  == size(Y) ||
-        throw(DimensionMismatch(
-            "Encountered two objects with sizes $(size(X)) and "*
-            "$(size(Y)) which needed to match but don't. "))
-    return nothing
-end
-
 """
     check_same_nrows(X, Y)
 

test/composition/learning_networks/replace.jl

@@ -33,8 +33,6 @@ zhat = inverse_transform(standM, uhat)
 yhat = exp(zhat)
 enode = @node mae(ys, yhat)
 
-_header(accel) =
-
 @testset "replace()  method; $(typeof(accel))" for accel in (CPU1(), CPUThreads())
 
     fit!(yhat, verbosity=0, acceleration=accel)
@@ -50,15 +48,12 @@ _header(accel) =
     knn2 = deepcopy(knn)
 
     # duplicate the network with `yhat` as glb:
-    yhat_clone = @test_logs(
-        (:warn, r"No replacement"),
-        replace(
-            yhat,
-            hot=>hot2,
-            knn=>knn2,
-            ys=>source(42);
-            copy_models_deeply=false,
-        ),
+    yhat_clone = replace(
+        yhat,
+        hot=>hot2,
+        knn=>knn2,
+        ys=>source(42);
+        copy_unspecified_deeply=false,
     )
 
     # test models and sources duplicated correctly:
@@ -79,16 +74,13 @@ _header(accel) =
     @test all(isempty, sources(yhat_ser))
 
     # duplicate a signature:
-    signature = (predict=yhat, report=(mae=enode,)) |> MLJBase.signature
-    signature_clone = @test_logs(
-        (:warn, r"No replacement"),
-        replace(
-            signature,
-            hot=>hot2,
-            knn=>knn2,
-            ys=>source(42);
-            copy_models_deeply=false,
-        )
+    signature = (predict=yhat, report=(mae=enode,)) |> MLJBase.Signature
+    signature_clone = replace(
+        signature,
+        hot=>hot2,
+        knn=>knn2,
+        ys=>source(2*y);
+        copy_unspecified_deeply=false,
     )
     glb_node = glb(signature_clone)
     models_clone = MLJBase.models(glb_node)
@@ -97,28 +89,20 @@ _header(accel) =
     @test models_clone[3] === hot2
     sources_clone = sources(glb_node)
     @test sources_clone[1]() == X
-    @test sources_clone[2]() === 42
+    @test sources_clone[2]() == 2*y
+
+    # warning thrown
+    @test_logs(
+        (:warn, r"No replacement"),
+        replace(
+            signature,
+            hot=>hot2,
+            knn=>knn2,
+            ys=>source(2*y);
+        ),
+    )
 
-    # duplicate a learning network machine:
-    mach  = machine(Deterministic(), Xs, ys;
-                    predict=yhat,
-                    report=(mae=enode,))
-    mach2 = replace(mach, hot=>hot2, knn=>knn2,
-                    ys=>source(ys.data);
-                    empty_unspecified_sources=true)
-    ss = sources(glb(mach2))
-    @test isempty(ss[1])
-    mach2 = @test_logs((:warn, r"No replacement"),
-                       replace(mach, hot=>hot2, knn=>knn2,
-                               ys=>source(ys.data)))
-    yhat2 = mach2.fitresult.predict
-    fit!(mach, verbosity=0)
-    fit!(mach2, verbosity=0)
-    @test predict(mach, X) ≈ predict(mach2, X)
-    @test report(mach).mae ≈ report(mach2).mae
-
-    @test mach2.args[1]() == Xs()
-    @test mach2.args[2]() == ys()
+    yhat2 = MLJBase.operation_nodes(signature_clone).predict
 
 
     ## EXTRA TESTS FOR TRAINING SEQUENCE
@@ -141,9 +125,7 @@ _header(accel) =
 
     @test length(MLJBase.machines(yhat)) == length(MLJBase.machines(yhat2))
     @test MLJBase.models(yhat) == MLJBase.models(yhat2)
-    @test sources(yhat) == sources(yhat2)
-    @test MLJBase.tree(yhat) == MLJBase.tree(yhat2)
-    @test yhat() ≈ yhat2()
+    @test 2yhat() ≈ yhat2()
 
     # this change should trigger retraining of all machines except the
     # univariate standardizer:
@@ -159,7 +141,6 @@ _header(accel) =
                          (:train, oakM2), (:train, knnM2)])
 end
 
-
 end # module
 
 true

test/utilities.jl

@@ -205,6 +205,16 @@ MLJBase.target_scitype(::Type{<:DRegressor2}) =
 
 @test MLJBase.guess_model_target_observation_scitype(DRegressor2()) == Continuous
 
+@testset "pretty" begin
+    X = (x=fill(1, 3), y=fill(2, 3))
+    io = IOBuffer()
+    pretty(X)
+    pretty(io, X)
+    str = take!(io) |> String
+    @test contains(str, "x")
+    @test contains(str, "y")
+    @test contains(str, "│")
+end
 
 end # module
 true