def test_ga_roulette_selection():
new_ga = ga.GA()
fitness = [10, 20, 30, 40, 50]
idx = new_ga._roulette_selection(len(fitness), fitness)
assert len(idx) == 4
def test_ga_update():
def square(x):
return np.sum(x**2)
new_ga = ga.GA()
search_space = search.SearchSpace(n_agents=10, n_variables=2,
lower_bound=[1, 1], upper_bound=[10, 10])
new_ga.update(search_space, square)
def test_ga_mutation():
search_space = search.SearchSpace(n_agents=10, n_variables=2,
lower_bound=[1, 1], upper_bound=[10, 10])
new_ga = ga.GA()
alpha, beta = new_ga._mutation(
search_space.agents[0], search_space.agents[1])
assert type(alpha).__name__ == 'Agent'
assert type(beta).__name__ == 'Agent'
def test_ga_hyperparams():
hyperparams = {
'p_selection': 0.75,
'p_mutation': 0.25,
'p_crossover': 0.5,
}
new_ga = ga.GA(hyperparams=hyperparams)
assert new_ga.p_selection == 0.75
assert new_ga.p_mutation == 0.25
assert new_ga.p_crossover == 0.5
def test_ga_params():
params = {
"p_selection": 0.75,
"p_mutation": 0.25,
"p_crossover": 0.5,
}
new_ga = ga.GA(params=params)
assert new_ga.p_selection == 0.75
assert new_ga.p_mutation == 0.25
assert new_ga.p_crossover == 0.5
def test_ga_update():
def square(x):
return np.sum(x**2)
new_function = function.Function(pointer=square)
new_ga = ga.GA()
search_space = search.SearchSpace(n_agents=10,
n_iterations=10,
n_variables=2,
lower_bound=[1, 1],
upper_bound=[10, 10])
new_ga._evaluate(search_space, new_function)
new_ga._update(search_space.agents, new_function)
assert search_space.agents[0].position[0] != 0
def test_ga_hyperparams_setter():
new_ga = ga.GA()
try:
new_ga.p_selection = 'a'
except:
new_ga.p_selection = 0.75
try:
new_ga.p_selection = -1
except:
new_ga.p_selection = 0.75
assert new_ga.p_selection == 0.75
try:
new_ga.p_mutation = 'b'
except:
new_ga.p_mutation = 0.25
try:
new_ga.p_mutation = -1
except:
new_ga.p_mutation = 0.25
assert new_ga.p_mutation == 0.25
try:
new_ga.p_crossover = 'c'
except:
new_ga.p_crossover = 0.5
try:
new_ga.p_crossover = -1
except:
new_ga.p_crossover = 0.5
assert new_ga.p_crossover == 0.5
def test_ga_params_setter():
new_ga = ga.GA()
try:
new_ga.p_selection = "a"
except:
new_ga.p_selection = 0.75
try:
new_ga.p_selection = -1
except:
new_ga.p_selection = 0.75
assert new_ga.p_selection == 0.75
try:
new_ga.p_mutation = "b"
except:
new_ga.p_mutation = 0.25
try:
new_ga.p_mutation = -1
except:
new_ga.p_mutation = 0.25
assert new_ga.p_mutation == 0.25
try:
new_ga.p_crossover = "c"
except:
new_ga.p_crossover = 0.5
try:
new_ga.p_crossover = -1
except:
new_ga.p_crossover = 0.5
assert new_ga.p_crossover == 0.5
def test_ga_run():
def square(x):
return np.sum(x**2)
def hook(optimizer, space, function):
return
new_function = function.Function(pointer=square)
new_ga = ga.GA()
search_space = search.SearchSpace(n_agents=10,
n_iterations=30,
n_variables=2,
lower_bound=[0, 0],
upper_bound=[10, 10])
history = new_ga.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 ga failed to converge.'
def test_ga_build():
new_ga = ga.GA()
assert new_ga.built == True
