def test_wwo_update_wave_length():
search_space = search.SearchSpace(n_agents=50,
n_variables=2,
lower_bound=[0, 0],
upper_bound=[10, 10])
new_wwo = wwo.WWO()
new_wwo.compile(search_space)
new_wwo._update_wave_length(search_space.agents)
def test_wwo_params():
params = {"h_max": 5, "alpha": 1.001, "beta": 0.001, "k_max": 1}
new_wwo = wwo.WWO(params=params)
assert new_wwo.h_max == 5
assert new_wwo.alpha == 1.001
assert new_wwo.beta == 0.001
assert new_wwo.k_max == 1
def test_wwo_params():
params = {'h_max': 5, 'alpha': 1.001, 'beta': 0.001, 'k_max': 1}
new_wwo = wwo.WWO(params=params)
assert new_wwo.h_max == 5
assert new_wwo.alpha == 1.001
assert new_wwo.beta == 0.001
assert new_wwo.k_max == 1
def test_wwo_params_setter():
new_wwo = wwo.WWO()
try:
new_wwo.h_max = "a"
except:
new_wwo.h_max = 5
try:
new_wwo.h_max = -1
except:
new_wwo.h_max = 5
assert new_wwo.h_max == 5
try:
new_wwo.alpha = "b"
except:
new_wwo.alpha = 1.001
try:
new_wwo.alpha = -1
except:
new_wwo.alpha = 1.001
assert new_wwo.alpha == 1.001
try:
new_wwo.beta = "c"
except:
new_wwo.beta = 0.001
try:
new_wwo.beta = -1
except:
new_wwo.beta = 0.001
assert new_wwo.beta == 0.001
try:
new_wwo.k_max = "d"
except:
new_wwo.k_max = 1
try:
new_wwo.k_max = -1
except:
new_wwo.k_max = 1
assert new_wwo.k_max == 1
def test_wwo_update():
def square(x):
return np.sum(x**2)
search_space = search.SearchSpace(n_agents=50,
n_variables=2,
lower_bound=[0, 0],
upper_bound=[10, 10])
new_wwo = wwo.WWO()
new_wwo.compile(search_space)
new_wwo.update(search_space, square)
new_wwo.update(search_space, square)
def test_wwo_hyperparams_setter():
new_wwo = wwo.WWO()
try:
new_wwo.h_max = 'a'
except:
new_wwo.h_max = 5
try:
new_wwo.h_max = -1
except:
new_wwo.h_max = 5
assert new_wwo.h_max == 5
try:
new_wwo.alpha = 'b'
except:
new_wwo.alpha = 1.001
try:
new_wwo.alpha = -1
except:
new_wwo.alpha = 1.001
assert new_wwo.alpha == 1.001
try:
new_wwo.beta = 'c'
except:
new_wwo.beta = 0.001
try:
new_wwo.beta = -1
except:
new_wwo.beta = 0.001
assert new_wwo.beta == 0.001
try:
new_wwo.k_max = 'd'
except:
new_wwo.k_max = 1
try:
new_wwo.k_max = -1
except:
new_wwo.k_max = 1
assert new_wwo.k_max == 1
def test_wwo_propagate_wave():
def square(x):
return np.sum(x**2)
search_space = search.SearchSpace(n_agents=50,
n_variables=2,
lower_bound=[0, 0],
upper_bound=[10, 10])
new_wwo = wwo.WWO()
new_wwo.compile(search_space)
wave = new_wwo._propagate_wave(search_space.agents[0], square, 0)
assert type(wave).__name__ == "Agent"
def test_wwo_refract_wave():
def square(x):
return np.sum(x**2)
search_space = search.SearchSpace(n_agents=50,
n_variables=2,
lower_bound=[0, 0],
upper_bound=[10, 10])
new_wwo = wwo.WWO()
new_wwo.compile(search_space)
height, length = new_wwo._refract_wave(search_space.agents[0],
search_space.best_agent, square, 0)
assert height == 5
assert length != 0
def test_wwo_compile():
search_space = search.SearchSpace(n_agents=50,
n_variables=2,
lower_bound=[0, 0],
upper_bound=[10, 10])
new_wwo = wwo.WWO()
new_wwo.compile(search_space)
try:
new_wwo.height = 1
except:
new_wwo.height = np.array([1])
assert new_wwo.height == np.array([1])
try:
new_wwo.length = 1
except:
new_wwo.length = np.array([1])
assert new_wwo.length == np.array([1])
def test_wwo_run():
def square(x):
return np.sum(x**2)
def hook(optimizer, space, function):
return
new_function = function.Function(pointer=square)
new_wwo = wwo.WWO({'k_max': 20})
search_space = search.SearchSpace(n_agents=10, n_iterations=100,
n_variables=2, lower_bound=[0, 0],
upper_bound=[10, 10])
history = new_wwo.run(search_space, new_function, pre_evaluation=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 wwo failed to converge.'
def test_wwo_build():
new_wwo = wwo.WWO()
assert new_wwo.built == True