Adjust Michalewicz benchmark

Author: AVHopp

.github/workflows/manual_benchmark.yml

@@ -52,8 +52,8 @@ on:
         required: false
         default: false
         type: boolean
-      michalewicz_tl_noise:
-        description: "michalewicz_tl_noise_benchmark"
+      michalewicz_tl_continuous:
+        description: "michalewicz_tl_continuous_benchmark"
         required: false
         default: false
         type: boolean

benchmarks/domains/__init__.py

@@ -24,7 +24,7 @@
     hartmann_tl_3_30_negate_15_benchmark,
 )
 from benchmarks.domains.transfer_learning.michalewicz.michalewicz_tl_continuous import (
-    michalewicz_tl_noise_benchmark,
+    michalewicz_tl_continuous_benchmark,
 )
 
 BENCHMARKS: list[Benchmark] = [
@@ -36,7 +36,7 @@
     easom_tl_100_negate_noise5_benchmark,
     hartmann_tl_3_30_negate_15_benchmark,
     hartmann_tl_3_20_15_benchmark,
-    michalewicz_tl_noise_benchmark,
+    michalewicz_tl_continuous_benchmark,
 ]
 
 

benchmarks/domains/transfer_learning/michalewicz/michalewicz_tl_continuous.py

@@ -17,6 +17,7 @@
 from baybe.campaign import Campaign
 from baybe.objectives import SingleTargetObjective
 from baybe.parameters import NumericalContinuousParameter, TaskParameter
+from baybe.parameters.base import Parameter
 from baybe.searchspace import SearchSpace
 from baybe.simulation import simulate_scenarios
 from baybe.targets import NumericalTarget
@@ -25,7 +26,7 @@
 
 def create_searchspace(use_task_parameter: bool) -> SearchSpace:
     """Create search space for the benchmark."""
-    params = [
+    params: list[Parameter] = [
         NumericalContinuousParameter(
             name=f"x{k}",
             bounds=(0, math.pi),
@@ -100,7 +101,7 @@ def create_initial_data(
     return df
 
 
-def michalewicz_tl_noise(settings: ConvergenceBenchmarkSettings) -> pd.DataFrame:
+def michalewicz_tl_continuous(settings: ConvergenceBenchmarkSettings) -> pd.DataFrame:
     """Benchmark function comparing TL and non-TL campaigns.
 
     Inputs:
@@ -130,7 +131,7 @@ def michalewicz_tl_noise(settings: ConvergenceBenchmarkSettings) -> pd.DataFrame
     for p in [1, 10, 25, 50, 100]:
         results.append(
             simulate_scenarios(
-                {f"{p}": campaign_tl},
+                {f"{p}": campaign_tl, f"{p}_naive": campaign_nontl},
                 lambda x: wrap_function(
                     functions["Target_Function"], "Target_Function", x
                 ),
@@ -163,8 +164,8 @@ def michalewicz_tl_noise(settings: ConvergenceBenchmarkSettings) -> pd.DataFrame
     n_mc_iterations=30,
 )
 
-michalewicz_tl_noise_benchmark = ConvergenceBenchmark(
-    function=michalewicz_tl_noise,
+michalewicz_tl_continuous_benchmark = ConvergenceBenchmark(
+    function=michalewicz_tl_continuous,
     optimal_target_values={"Target": 4.687658},
     settings=benchmark_config,
 )