def test_mutate(self):
def get_space():
space = HyperSpace()
with space.as_default():
id1 = Identity(p1=Int(0, 10), p2=Choice(['a', 'b']))
id2 = Identity(p3=Real(0., 1.), p4=Bool())(id1)
return space
# population = Population(optimize_direction=OptimizeDirection.Maximize)
population = Population(optimize_direction='max')
space1 = get_space()
space1.random_sample()
assert space1.all_assigned
space2 = get_space()
assert not space2.all_assigned
new_space = population.mutate(space1, space2)
pv1 = list(space1.get_assigned_param_values().values())
pv2 = list(space2.get_assigned_param_values().values())
assert space2.all_assigned
assert new_space.all_assigned
assert np.sum([v1 != v2 for v1, v2 in zip(pv1, pv2)]) == 1
def test_eliminate(self):
population = Population(optimize_direction=OptimizeDirection.Maximize)
population.append('a', 4)
population.append('b', 3)
population.append('c', 2)
population.append('d', 1)
population.append('e', 0)
population.append('f', 5)
population.append('g', 6)
population.append('h', 7)
population.append('i', 8)
population.append('j', 9)
eliminates = population.eliminate(2, regularized=True)
assert eliminates[0].space_sample == 'a' and eliminates[
1].space_sample == 'b'
eliminates = population.eliminate(2, regularized=False)
assert eliminates[0].space_sample == 'e' and eliminates[
1].space_sample == 'd'
def test_population(self):
population = Population(optimize_direction=OptimizeDirection.Maximize)
population.append('a', 0)
population.append('b', 1)
population.append('c', 2)
population.append('d', 3)
population.append('e', 4)
population.append('f', 5)
population.append('g', 6)
population.append('h', 7)
population.append('i', 8)
population.append('i', 9)
b1 = population.sample_best(25, np.random.RandomState(9527))
assert b1.reward == 8
population = Population(optimize_direction=OptimizeDirection.Minimize)
population.append('a', 0)
population.append('b', 1)
population.append('c', 2)
population.append('d', 3)
population.append('e', 4)
population.append('f', 5)
population.append('g', 6)
population.append('h', 7)
population.append('i', 8)
population.append('i', 9)
b2 = population.sample_best(25, np.random.RandomState(9527))
assert b2.reward == 0