Merge pull request #3 from JuliaAI/coverage

Make tests more verbose to increase test coverage

Author: ablaom

.github/codecov.yml

@@ -1,5 +1,6 @@
 coverage:
-  status:
+    ignore:
+       - "src/learners/gradient_descent.jl"
     project:
       default:
         threshold: 0.5%

test/learners/dimension_reduction.jl

@@ -11,7 +11,7 @@ U, Vt = r.U, r.Vt
 X = U*diagm([1, 2, 3, 0.01, 0.01])*Vt
 
 learner = LearnTestAPI.TruncatedSVD(codim=2)
-@testapi learner X verbosity=0
+@testapi learner X verbosity=1
 
 @testset "extra test for truncated SVD" begin
     model = @test_logs(

test/learners/ensembling.jl

@@ -25,7 +25,7 @@ Xtest = Tables.subset(X, test)
 rng = StableRNG(123)
 atom = LearnTestAPI.Ridge()
 learner = LearnTestAPI.Ensemble(atom; n=4, rng)
-@testapi learner data verbosity=0
+@testapi learner data verbosity=1
 
 @testset "extra tests for ensemble" begin
     @test LearnAPI.clone(learner) == learner

test/learners/regression.jl

@@ -20,7 +20,7 @@ data = (X, y)
 # # RIDGE
 
 learner = LearnTestAPI.Ridge(lambda=0.5)
-@testapi learner data verbosity=0
+@testapi learner data verbosity=1
 
 @testset "extra tests for ridge regression" begin
     @test :(LearnAPI.obs) in LearnAPI.functions(learner)
@@ -82,7 +82,7 @@ end
 # # BABY RIDGE
 
 learner = LearnTestAPI.BabyRidge(lambda=0.5)
-@testapi learner data verbosity=0
+@testapi learner data verbosity=1
 
 @testset "extra tests for baby ridge" begin
     model = fit(learner, Tables.subset(X, train), y[train]; verbosity=0)

test/learners/static_algorithms.jl

@@ -9,7 +9,7 @@ import DataFrames
 
 learner = LearnTestAPI.Selector(names=[:x, :w])
 X = DataFrames.DataFrame(rand(3, 4), [:x, :y, :z, :w])
-@testapi learner X verbosity=0
+@testapi learner X verbosity=1
 
 @testset "test a static transformer" begin
     model = fit(learner) # no data arguments!
@@ -23,7 +23,7 @@ end
 
 learner = LearnTestAPI.FancySelector(names=[:x, :w])
 X = DataFrames.DataFrame(rand(3, 4), [:x, :y, :z, :w])
-@testapi learner X verbosity=0
+@testapi learner X verbosity=1
 
 @testset "test a variation that reports byproducts" begin
     model = fit(learner) # no data arguments!