revert tests

Author: fkiraly

…peractive/tests/integrations/__init__.py …test_optimization_strategies/__init__.pysrc/hyperactive/tests/integrations/__init__.py renamed to tests/_local_test_optimization_strategies/__init__.py src/hyperactive/tests/integrations/__init__.py renamed to tests/_local_test_optimization_strategies/__init__.py

@@ -0,0 +1,110 @@
+from hyperactive.optimizers import (
+    HillClimbingOptimizer,
+    StochasticHillClimbingOptimizer,
+    RepulsingHillClimbingOptimizer,
+    SimulatedAnnealingOptimizer,
+    DownhillSimplexOptimizer,
+    RandomSearchOptimizer,
+    GridSearchOptimizer,
+    RandomRestartHillClimbingOptimizer,
+    RandomAnnealingOptimizer,
+    PowellsMethod,
+    PatternSearch,
+    ParallelTemperingOptimizer,
+    ParticleSwarmOptimizer,
+    SpiralOptimization,
+    EvolutionStrategyOptimizer,
+    BayesianOptimizer,
+    LipschitzOptimizer,
+    DirectAlgorithm,
+    TreeStructuredParzenEstimators,
+    ForestOptimizer,
+)
+
+
+optimizers = (
+    "Optimizer",
+    [
+        (HillClimbingOptimizer),
+        (StochasticHillClimbingOptimizer),
+        (RepulsingHillClimbingOptimizer),
+        (SimulatedAnnealingOptimizer),
+        (DownhillSimplexOptimizer),
+        (RandomSearchOptimizer),
+        (GridSearchOptimizer),
+        (RandomRestartHillClimbingOptimizer),
+        (RandomAnnealingOptimizer),
+        (PowellsMethod),
+        (PatternSearch),
+        (ParallelTemperingOptimizer),
+        (ParticleSwarmOptimizer),
+        (SpiralOptimization),
+        (EvolutionStrategyOptimizer),
+        (BayesianOptimizer),
+        (LipschitzOptimizer),
+        (DirectAlgorithm),
+        (TreeStructuredParzenEstimators),
+        (ForestOptimizer),
+    ],
+)
+
+
+optimizers_strat = (
+    "Optimizer_strat",
+    [
+        (HillClimbingOptimizer),
+        (StochasticHillClimbingOptimizer),
+        (RepulsingHillClimbingOptimizer),
+        (SimulatedAnnealingOptimizer),
+        (DownhillSimplexOptimizer),
+        (RandomSearchOptimizer),
+        (GridSearchOptimizer),
+        (RandomRestartHillClimbingOptimizer),
+        (RandomAnnealingOptimizer),
+        (PowellsMethod),
+        (PatternSearch),
+        (ParallelTemperingOptimizer),
+        (ParticleSwarmOptimizer),
+        (SpiralOptimization),
+        (EvolutionStrategyOptimizer),
+        (BayesianOptimizer),
+        (LipschitzOptimizer),
+        (DirectAlgorithm),
+        (TreeStructuredParzenEstimators),
+        (ForestOptimizer),
+    ],
+)
+
+
+optimizers_non_smbo = (
+    "Optimizer_non_smbo",
+    [
+        (HillClimbingOptimizer),
+        (StochasticHillClimbingOptimizer),
+        (RepulsingHillClimbingOptimizer),
+        (SimulatedAnnealingOptimizer),
+        (DownhillSimplexOptimizer),
+        (RandomSearchOptimizer),
+        (GridSearchOptimizer),
+        (RandomRestartHillClimbingOptimizer),
+        (RandomAnnealingOptimizer),
+        (PowellsMethod),
+        (PatternSearch),
+        (ParallelTemperingOptimizer),
+        (ParticleSwarmOptimizer),
+        (SpiralOptimization),
+        (EvolutionStrategyOptimizer),
+    ],
+)
+
+
+optimizers_smbo = (
+    "Optimizer_smbo",
+    [
+        (BayesianOptimizer),
+        (LipschitzOptimizer),
+        (DirectAlgorithm),
+        (TreeStructuredParzenEstimators),
+        (ForestOptimizer),
+    ],
+)

tests/_local_test_optimization_strategies/_parametrize.py

@@ -0,0 +1,155 @@
+import time
+import pytest
+import numpy as np
+
+
+from hyperactive import Hyperactive
+from hyperactive.optimizers.strategies import CustomOptimizationStrategy
+from hyperactive.optimizers import GridSearchOptimizer
+
+from ._parametrize import optimizers_non_smbo
+
+
+def objective_function(opt):
+    time.sleep(0.01)
+    score = -(opt["x1"] * opt["x1"])
+    return score
+
+
+search_space = {
+    "x1": list(np.arange(0, 100, 1)),
+}
+
+
+def test_memory_Warm_start_0():
+    optimizer1 = GridSearchOptimizer()
+    optimizer2 = GridSearchOptimizer()
+
+    opt_strat = CustomOptimizationStrategy()
+    opt_strat.add_optimizer(optimizer1, duration=0.2)
+    opt_strat.add_optimizer(optimizer2, duration=0.8)
+
+    n_iter = 1000
+
+    c_time = time.time()
+
+    hyper = Hyperactive()
+    hyper.add_search(
+        objective_function,
+        search_space,
+        optimizer=opt_strat,
+        n_iter=n_iter,
+        memory=True,
+    )
+    hyper.run()
+
+    d_time = time.time() - c_time
+
+    search_data = hyper.search_data(objective_function)
+    
+    optimizer1 = hyper.opt_pros[0].optimizer_setup_l[0]["optimizer"]
+    optimizer2 = hyper.opt_pros[0].optimizer_setup_l[1]["optimizer"]
+
+    assert len(search_data) == n_iter
+
+    assert len(optimizer1.search_data) == 200
+    assert len(optimizer2.search_data) == 800
+
+    assert optimizer1.best_score <= optimizer2.best_score
+
+    print("\n d_time", d_time)
+
+    assert d_time < 3
+
+
+def test_memory_Warm_start_1():
+    optimizer1 = GridSearchOptimizer()
+    optimizer2 = GridSearchOptimizer()
+
+    opt_strat = CustomOptimizationStrategy()
+    opt_strat.add_optimizer(optimizer1, duration=0.2)
+    opt_strat.add_optimizer(optimizer2, duration=0.8)
+
+    n_iter = 100
+
+    search_space = {
+        "x1": list(np.arange(0, 1, 1)),
+    }
+
+    c_time = time.time()
+
+    hyper = Hyperactive()
+    hyper.add_search(
+        objective_function,
+        search_space,
+        optimizer=opt_strat,
+        n_iter=n_iter,
+        memory=False,
+    )
+    hyper.run()
+
+    d_time = time.time() - c_time
+
+    search_data = hyper.search_data(objective_function)
+
+    optimizer1 = hyper.opt_pros[0].optimizer_setup_l[0]["optimizer"]
+    optimizer2 = hyper.opt_pros[0].optimizer_setup_l[1]["optimizer"]
+
+    assert len(search_data) == n_iter
+
+    assert len(optimizer1.search_data) == 20
+    assert len(optimizer2.search_data) == 80
+
+    assert optimizer1.best_score <= optimizer2.best_score
+
+    print("\n d_time", d_time)
+
+    assert d_time > 0.95
+
+
+
+@pytest.mark.parametrize(*optimizers_non_smbo)
+def test_memory_Warm_start_2(Optimizer_non_smbo):
+    optimizer1 = GridSearchOptimizer()
+    optimizer2 = Optimizer_non_smbo()
+
+    opt_strat = CustomOptimizationStrategy()
+    opt_strat.add_optimizer(optimizer1, duration=0.5)
+    opt_strat.add_optimizer(optimizer2, duration=0.5)
+
+    search_space = {
+        "x1": list(np.arange(0, 50, 1)),
+    }
+
+    n_iter = 100
+
+    c_time = time.time()
+
+    hyper = Hyperactive()
+    hyper.add_search(
+        objective_function,
+        search_space,
+        optimizer=opt_strat,
+        n_iter=n_iter,
+        memory=True,
+    )
+    hyper.run()
+
+    d_time = time.time() - c_time
+
+    search_data = hyper.search_data(objective_function)
+
+    optimizer1 = hyper.opt_pros[0].optimizer_setup_l[0]["optimizer"]
+    optimizer2 = hyper.opt_pros[0].optimizer_setup_l[1]["optimizer"]
+
+    assert len(search_data) == n_iter
+
+    assert len(optimizer1.search_data) == 50
+    assert len(optimizer2.search_data) == 50
+
+    assert optimizer1.best_score <= optimizer2.best_score
+
+    print("\n d_time", d_time)
+
+    assert d_time < 0.9
+    

tests/_local_test_optimization_strategies/_test_memory_warm_start.py

@@ -0,0 +1,55 @@
+import time
+import pytest
+import numpy as np
+
+
+from hyperactive import Hyperactive
+from hyperactive.optimizers.strategies import CustomOptimizationStrategy
+from hyperactive.optimizers import GridSearchOptimizer
+
+from ._parametrize import optimizers_smbo
+
+
+def objective_function(opt):
+    time.sleep(0.01)
+    score = -(opt["x1"] * opt["x1"])
+    return score
+
+
+search_space = {
+    "x1": list(np.arange(0, 100, 1)),
+}
+
+
+@pytest.mark.parametrize(*optimizers_smbo)
+def test_memory_Warm_start_smbo_0(Optimizer_smbo):
+    optimizer1 = GridSearchOptimizer()
+    optimizer2 = Optimizer_smbo()
+
+    opt_strat = CustomOptimizationStrategy()
+    opt_strat.add_optimizer(optimizer1, duration=0.8)
+    opt_strat.add_optimizer(optimizer2, duration=0.2)
+
+    n_iter = 100
+
+    hyper = Hyperactive()
+    hyper.add_search(
+        objective_function,
+        search_space,
+        optimizer=opt_strat,
+        n_iter=n_iter,
+        memory=True,
+    )
+    hyper.run()
+
+    search_data = hyper.search_data(objective_function)
+
+    optimizer1 = hyper.opt_pros[0].optimizer_setup_l[0]["optimizer"]
+    optimizer2 = hyper.opt_pros[0].optimizer_setup_l[1]["optimizer"]
+
+    assert len(search_data) == n_iter
+
+    assert len(optimizer1.search_data) == 80
+    assert len(optimizer2.search_data) == 20
+
+    assert optimizer1.best_score <= optimizer2.best_score