Confirm that the problem occurs on 2 datasets

rikhuijzer · Jun 21, 2023 · 31d56f3 · 31d56f3
1 parent 365137d
commit 31d56f3
Show file tree

Hide file tree

Showing 2 changed files with 21 additions and 3 deletions.
diff --git a/docs/Project.toml b/docs/Project.toml
@@ -9,6 +9,7 @@ Documenter = "e30172f5-a6a5-5a46-863b-614d45cd2de4"
 LightGBM = "7acf609c-83a4-11e9-1ffb-b912bcd3b04a"
 LiveServer = "16fef848-5104-11e9-1b77-fb7a48bbb589"
 MLJ = "add582a8-e3ab-11e8-2d5e-e98b27df1bc7"
+MLJBase = "a7f614a8-145f-11e9-1d2a-a57a1082229d"
 MLJDecisionTreeInterface = "c6f25543-311c-4c74-83dc-3ea6d1015661"
 Pkg = "44cfe95a-1eb2-52ea-b672-e2afdf69b78f"
 PlutoStaticHTML = "359b1769-a58e-495b-9770-312e911026ad"
@@ -26,6 +27,7 @@ DecisionTree = "0.12"
 Documenter = "0.27"
 LiveServer = "1"
 MLJ = "0.19"
+MLJBase = "0.21"
 MLJDecisionTreeInterface = "0.4"
 PlutoStaticHTML = "6"
 PlutoUI = "0.7"

diff --git a/test/mlj.jl b/test/mlj.jl
@@ -23,7 +23,10 @@ datasets = Dict{String,Tuple}(
             y = df.survival
             (X, y)
         end,
-        "boston" => boston()
+        "boston" => boston(),
+        "make_regression" => let
+             make_regression(200, 3; noise=0.0, sparse=0.0, outliers=0.0)
+         end
     )
 
 function _score(e::PerformanceEvaluation)
@@ -200,8 +203,21 @@ preds = predict(rulesmach)
 @test preds isa Vector{Float64}
 # @test 0.95 < rsq(preds, y)
 
-let
-    # e = _evaluate_baseline!(results, "boston")
+emr = let
+    hyper = (;)
+    measure = rsq
+    _evaluate!(results, "make_regression", LGBMRegressor, hyper; measure)
+    _evaluate!(results, "make_regression", LinearRegressor, hyper; measure)
+    _evaluate!(results, "make_regression", StableForestRegressor, hyper; measure)
+
+    hyper = (; rng=_rng(), n_trees=1_500, max_rules=500)
+    _evaluate!(results, "make_regression", StableRulesRegressor, hyper; measure)
+
+    hyper = (; rng=_rng(), n_trees=1_500, max_rules=30)
+    _evaluate!(results, "make_regression", StableRulesRegressor, hyper; measure)
+
+    hyper = (; rng=_rng(), n_trees=1_500, max_rules=10)
+    _evaluate!(results, "make_regression", StableRulesRegressor, hyper; measure)
 end
 
 el = let