def test_abc_run():
def square(x):
return np.sum(x**2)
def hook(optimizer, space, function):
return
new_function = function.Function(pointer=square)
hyperparams = {'n_trials': 1}
new_abc = abc.ABC(hyperparams=hyperparams)
search_space = search.SearchSpace(n_agents=10,
n_iterations=100,
n_variables=2,
lower_bound=[0, 0],
upper_bound=[10, 10])
history = new_abc.run(search_space, new_function, pre_evaluation_hook=hook)
assert len(history.agents) > 0
assert len(history.best_agent) > 0
best_fitness = history.best_agent[-1][1]
assert best_fitness <= constants.TEST_EPSILON, 'The algorithm abc failed to converge.'
def test_abc_hyperparams_setter():
new_abc = abc.ABC()
try:
new_abc.n_trials = 0.0
except:
new_abc.n_trials = 10
try:
new_abc.n_trials = 0
except:
new_abc.n_trials = 10
assert new_abc.n_trials == 10
def test_abc_run():
def square(x):
return np.sum(x**2)
new_function = function.Function(pointer=square)
hyperparams = {'n_trials': 10}
new_abc = abc.ABC(hyperparams=hyperparams)
search_space = search.SearchSpace(n_agents=2,
n_iterations=100,
n_variables=2,
lower_bound=[0, 0],
upper_bound=[10, 10])
history = new_abc.run(search_space, new_function)
assert len(history.agents) > 0
assert len(history.best_agent) > 0
def test_abc_build():
new_abc = abc.ABC()
assert new_abc.built == True
def test_abc_hyperparams_setter():
new_abc = abc.ABC()
new_abc.n_trials = 10
assert new_abc.n_trials == 10
def test_abc_hyperparams():
hyperparams = {'n_trials': 5}
new_abc = abc.ABC(hyperparams=hyperparams)
assert new_abc.n_trials == 5