def test_wca_hyperparams_setter():
new_wca = wca.WCA()
try:
new_wca.nsr = 0.0
except:
new_wca.nsr = 10
try:
new_wca.nsr = 0
except:
new_wca.nsr = 10
assert new_wca.nsr == 10
try:
new_wca.d_max = 'a'
except:
new_wca.d_max = 0.1
try:
new_wca.d_max = -1
except:
new_wca.d_max = 0.1
assert new_wca.d_max == 0.1
def test_wca_hyperparams_setter():
new_wca = wca.WCA()
new_wca.nsr = 10
assert new_wca.nsr == 10
new_wca.d_max = 0.1
assert new_wca.d_max == 0.1
def test_wca_hyperparams():
hyperparams = {'nsr': 5, 'd_max': 0.25}
new_wca = wca.WCA(hyperparams=hyperparams)
assert new_wca.nsr == 5
assert new_wca.d_max == 0.25
def test_wca_update_river():
new_wca = wca.WCA()
search_space = search.SearchSpace(n_agents=20, n_iterations=100,
n_variables=2, lower_bound=[0, 0],
upper_bound=[10, 10])
new_wca._update_river(search_space.agents, search_space.best_agent)
assert search_space.agents[1].position[0] != 0
def test_wca_flow_intensity():
new_wca = wca.WCA()
search_space = search.SearchSpace(n_agents=2, n_iterations=100,
n_variables=2, lower_bound=[0, 0],
upper_bound=[10, 10])
flows = new_wca._flow_intensity(search_space.agents)
assert flows.shape[0] == new_wca.nsr
def test_wca_update():
new_wca = wca.WCA()
search_space = search.SearchSpace(n_agents=20, n_iterations=100,
n_variables=2, lower_bound=[0, 0],
upper_bound=[10, 10])
flows = new_wca._flow_intensity(search_space.agents)
new_wca._update(search_space.agents, search_space.best_agent, flows)
assert search_space.agents[0].position[0] != 0
def test_wca_run():
def square(x):
return np.sum(x**2)
new_function = function.Function(pointer=square)
new_wca = wca.WCA()
search_space = search.SearchSpace(n_agents=20, n_iterations=10,
n_variables=2, lower_bound=[0, 0],
upper_bound=[10, 10])
history = new_wca.run(search_space, new_function)
assert len(history.agents) > 0
assert len(history.best_agent) > 0
def test_wca_run():
def square(x):
return np.sum(x**2)
def hook(optimizer, space, function):
return
new_function = function.Function(pointer=square)
new_wca = wca.WCA()
search_space = search.SearchSpace(n_agents=20,
n_iterations=10,
n_variables=2,
lower_bound=[0, 0],
upper_bound=[10, 10])
history = new_wca.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 wca failed to converge.'
def test_wca_build():
new_wca = wca.WCA()
assert new_wca.built == True