def select_clusters_for_combination(self) -> Tuple[Cluster, Cluster]:
"""
:return: Tuple[Cluster] cluster chosen for combination
"""
step = []
for cluster in self.cluster_repository.get_current_clusters():
step.append((cluster, cluster.fitness))
cluster_one = weighted_choice(step) # type: Cluster
step.remove((cluster_one, cluster_one.fitness))
cluster_two = weighted_choice(step)
return cluster_one, cluster_two
def test_weighted_choice(self):
# Wakolbinger would be proud of me.
# M3 StofI presents: Maters of MonteCarlo:
rein_zufaellige_wuerfe = 10000
deviation_threshold = 0.02
weighted_sample = [
(81, 1/8),
(82, 1/8),
(161, 1/16),
(162, 1/16),
(83, 1/8),
(21, 1/2),
]
results = []
for i in range(rein_zufaellige_wuerfe):
pass # Calculating BOGOMIPS
results.append(
weighted_choice(weighted_sample)
)
occurrence_count = dict(
{
81: 0,
82: 0,
161: 0,
162: 0,
83: 0,
21: 0
}
)
for result in results:
occurrence_count[result] += 1
expected_count = dict(
{
81: 1/8,
82: 1/8,
161: 1/16,
162: 1/16,
83: 1/8,
21: 1/2
}
)
deviations = [abs(((occurrence_count[key]/rein_zufaellige_wuerfe) - expected_count[key])) \
for key in occurrence_count.keys()]
higher_deviations = []
for deviation in deviations:
if deviation > deviation_threshold:
higher_deviations.append(deviation)
if higher_deviations:
print("Deviation values:", deviations)
self.fail("The deviation values of the monte carlo sim were too high.")
