def test_n_ary_crossover_bad_crossover_points():
""" Test assertions for having more crossover points than genome length """
pop = [Individual(np.array([0, 0])), Individual(np.array([1, 1]))]
i = ops.naive_cyclic_selection(pop)
with pytest.raises(RuntimeError):
new_pop = list(
itertools.islice(ops.n_ary_crossover(i, num_points=3), 2))
def test_n_ary_crossover_probability2():
"""If we perform uniform crossover with a probabilty of 1.0, then we should see genes swapped
by default with probability 0.2."""
N = 5000
observed_dist = {
'Unmodified': 0,
'Only left swapped': 0,
'Only right swapped': 0,
'Both swapped': 0
}
# Run crossover N times on a fixed pair of two-gene individuals
for i in range(N):
pop = [Individual(np.array([0, 0])), Individual(np.array([1, 1]))]
i = ops.naive_cyclic_selection(pop)
new_pop = list(
itertools.islice(ops.uniform_crossover(i, p_xover=1.0), 2))
# There are four possible outcomes, which we will count the occurence of
if np.all(new_pop[0].genome == [0, 0]) and np.all(
new_pop[1].genome == [1, 1]):
observed_dist['Unmodified'] += 1
elif np.all(new_pop[0].genome == [1, 0]) and np.all(
new_pop[1].genome == [0, 1]):
observed_dist['Only left swapped'] += 1
elif np.all(new_pop[0].genome == [0, 1]) and np.all(
new_pop[1].genome == [1, 0]):
observed_dist['Only right swapped'] += 1
elif np.all(new_pop[0].genome == [1, 1]) and np.all(
new_pop[1].genome == [0, 0]):
observed_dist['Both swapped'] += 1
else:
assert (False)
assert (N == sum(observed_dist.values()))
p = 0.01
p_swap = 0.2
# This is the count we expect to see of each combination
# Each locus swaps with p_swap.
expected_dist = {
'Unmodified': int((1 - p_swap) * (1 - p_swap) * N),
'Only left swapped': int(p_swap * (1 - p_swap) * N),
'Only right swapped': int((1 - p_swap) * p_swap * N),
'Both swapped': int(p_swap**2 * N)
}
# Use a χ-squared test to see if our experiment matches what we expect
assert (stat.stochastic_equals(expected_dist, observed_dist, p=p))
def test_n_ary_crossover():
"""If we crossover two individuals with two bits each, the children should either be swapped copies of their parents,
or they should exchange the second bit and keep the first bit unmodified."""
pop = [Individual(np.array([0, 0])), Individual(np.array([1, 1]))]
i = ops.naive_cyclic_selection(pop)
new_pop = list(itertools.islice(ops.n_ary_crossover(i, num_points=1), 2))
# Given that there are only two genes, one [0,0] and the other [1,1] and a single crossover point, and that the
# only two valid crossover points are 0 or 1, then there are two possible valid states for offspring with single
# point crossover.
assert np.all(pop[0].genome == [1, 1]) or np.all(pop[0].genome == [0, 1])
assert np.all(pop[1].genome == [0, 0]) or np.all(pop[1].genome == [1, 0])
def test_uniform_crossover():
pop = [Individual(np.array([0, 0])), Individual(np.array([1, 1]))]
# We need a cyclic generator because there are only two individuals in the population, and once the first two
# are selected for uniform crossover, the next two parents are selected and crossed over. The cyclic iterator
# ensures we just select the same two individuals again even though the yield statements in the uniform
# crossover operator are not invoked again.
i = ops.naive_cyclic_selection(pop)
# Do swap with 100% certainty, which will cause the two individuals' genomes to exchange values
new_pop = list(itertools.islice(ops.uniform_crossover(i, p_swap=1.0), 2))
assert np.all(new_pop[0].genome == [1, 1])
assert np.all(new_pop[1].genome == [0, 0])
# Note because we didn't clone the selected individuals, *the original population was changed*.
assert np.all(pop[0].genome == [1, 1])
assert np.all(pop[1].genome == [0, 0])
def test_naive_cyclic_selection():
""" Test of the naive deterministic cyclic selection """
pop = [
Individual(np.array([0, 0]), problem=MaxOnes()),
Individual(np.array([0, 1]), problem=MaxOnes())
]
# This selection operator will deterministically cycle through the
# given population
selector = ops.naive_cyclic_selection(pop)
selected = next(selector)
assert np.all(selected.genome == [0, 0])
selected = next(selector)
assert np.all(selected.genome == [0, 1])
# And now we cycle back to the first individual
selected = next(selector)
assert np.all(selected.genome == [0, 0])
def test_n_ary_crossover_probability():
"""If we perform crossover with a probabilty of 0.5, then the individuals will be unmodified 50% of the time."""
N = 1000
unmodified_count = 0
for i in range(N):
pop = [Individual([0, 0]), Individual([1, 1])]
i = ops.naive_cyclic_selection(pop)
new_pop = list(itertools.islice(ops.n_ary_crossover(i, p=0.5), 2))
if new_pop[0].genome == [0, 0] and new_pop[1].genome == [1, 1]:
unmodified_count += 1
p = 0.01
observed_dist = {
'Unmodified': unmodified_count,
'Modified': N - unmodified_count
}
assert (stat.equals_uniform(observed_dist, p=p))
def test_uniform_crossover_probability1():
"""If we perform uniform rossover with a probabilty of 0.0, then the individuals will always be unmodified.
This test calls the crossover opererator, which is stochastic, but we haven't marked it as part of the
stochastic test suite because there is no chance of a false failure (i.e. a test that fails even when
there is no fault in the code) in this case."""
N = 20
unmodified_count = 0
for i in range(N):
pop = [Individual(np.array([0, 0])), Individual(np.array([1, 1]))]
i = ops.naive_cyclic_selection(pop)
new_pop = list(
itertools.islice(ops.uniform_crossover(i, p_xover=0.0), 2))
if np.all(new_pop[0].genome == [0, 0]) and np.all(
new_pop[1].genome == [1, 1]):
unmodified_count += 1
assert (unmodified_count == N)