def test_connectivity_matrix():
"""
This test will ensure two things:
1. That all values of the connectivity matrix are 0 or 1, representing
nodes that are connected (1) or disconnected (0).
2. That the number of connections in a society using a very high
connectivity_probability is greater then a society using a very
low connectivity_probability.
"""
# TO DO FOR MITCHMAN
# First, go through the matrix and test to make sure all edges are an
# appropriate value.
# Instantiate constants
population_size = 50
connectivity_probability = 0.15
# Create society object.
society = Society(
population_size=population_size,
connectivity_probability=connectivity_probability,
)
# Loop through each entry and assert either 0 or 1.
for x in range(population_size):
for y in range(population_size):
assert (society.edge_matrix[x][y] == 0) or (society.edge_matrix[x][y] == 1)
# Next, create two society classes, one with connectivity probability 0.99
# and another with probability 0.01. Both will have larger population
# sizes to avoid fluke statistical generation. Assert sums of the two
# edge matrices.
# Set up constants
population_size = 1000
connectivity_probability_first = 0.99
connectivity_probability_second = 0.01
# Create both society objects
society_first = Society(
population_size=population_size,
connectivity_probability=connectivity_probability_first,
)
society_second = Society(
population_size=population_size,
connectivity_probability=connectivity_probability_second,
)
# Sum edges, and assert that the first has more than the second.
count_first = 0
count_second = 0
for x in range(population_size):
for y in range(population_size):
if society_first.edge_matrix[x][y] == 1:
count_first += 1
if society_second.edge_matrix[x][y] == 1:
count_second += 1
assert count_first > count_second
def demo_2():
# 遺伝子長
dimension = 30
# 集団数
individual_num = 300
# 世代数の最大
generation_loop = 1000
# 評価関数
evaluator = Rosenbrock()
# 親の選択と交叉
# 親の数は交叉方法に合わせて設定する
# 次世代に残す個体の選択
# Simplex, ルーレット選択
society = Society(
evaluator,
Simplex(dimension * 10),
RouletteSelector(dimension + 1),
JGG())
generator = Generator(10., -10., dimension)
society.generate_individuals(individual_num, generator)
# 初期個体の生成
society.generate_individuals(individual_num, generator)
# 画像データ表示用
best_value = []
iterate = []
generation_data = []
generation_data.append(get_generation_data(society.individuals))
best = None
for i in range(generation_loop):
society.change_generation()
# 画像データ表示用処理
best = society.get_best_individual()
iterate.append(i)
best_value.append(best.evaluate_value)
generation_data.append(get_generation_data(society.individuals))
print(i, best.evaluate_value)
# 最良評価値の推移を表示
plot_transition(iterate, best_value, generation_data)
def test_lawInstitution_fields():
# Verify, upon instantiation, that all fields of the law and institution
# classes are of the type they should be.
society = Society(population_size=10)
institution = Institution(society)
laws = institution._generate_laws(society)
assert isinstance(institution.get_affiliation(), float)
for law in laws:
for person_id in range(society.population_size):
assert isinstance(law.affected_persons[person_id], int)
assert isinstance(law.mean, float)
assert isinstance(law.law_type, str)
assert isinstance(law.actual_value, float)
def test_population():
"""
This test will simply ensure that a created Society object has a list of
Person objects, that is of length population_size.
"""
# Set up constants
population_size = 50
connectivity_probability = 0.15
# Create Society object.
society = Society(
population_size=population_size,
connectivity_probability=connectivity_probability,
)
# Assert the correct population.
assert len(society.person_vector) == population_size
# Assert each object in society.person_vector is a Person object
for person in society.person_vector:
assert type(person) == Person
def test_toy_model():
random.seed(42)
s = Society(episodes_per_day=5, encounter_size=2)
d = Disease(days_infectious=10, pr_transmission_per_day=0.2)
# seed size is the number of people in the population who we seed as being infected:
o = Outbreak(s,
d,
pop_size=1000,
seed_size=2,
n_days=ALL_TIME_DAYS,
population_type=Population,
person_type=Person)
# so, this is a village of 10000 people with 2 starting off infected
o.simulate()
assert o.recorder.story[90:95] == \
[[18.199999999999967, 0.619, 0.57, 0.0, 0.0, 0.0],
[18.399999999999967, 0.643, 0.592, 0.0, 0.0, 0.0],
[18.599999999999966, 0.656, 0.603, 0.0, 0.0, 0.0],
[18.799999999999965, 0.673, 0.618, 0.0, 0.0, 0.0],
[18.999999999999964, 0.69, 0.635, 0.0, 0.0, 0.0]]
def demo_3():
# 遺伝子長
dimension = 30
# 集団数
individual_num = 300
# 世代数の最大
generation_loop = 1000
# 評価関数
evaluator = Rosenbrock()
# 次世代に残す個体の選択
society_simplex = Society(
evaluator,
Simplex(dimension * 10),
RouletteSelector(dimension + 1),
JGG())
generator = Generator(10., -10., dimension)
society_blx = Society(
evaluator,
BLX_alpha(dimension * 10),
RouletteSelector(2),
JGG())
# 初期個体の生成
society_simplex.generate_individuals(individual_num, generator)
society_blx.generate_individuals(individual_num, generator)
# 画像データ表示用
simplex_data = []
blx_data = []
iterator_list = []
best_simplex = society_simplex.get_best_individual()
best_blx = society_blx.get_best_individual()
simplex_data.append(best_simplex.evaluate_value)
best_blx.append(best_blx.evaluate_value)
iterator_list.append()
for i in range(generation_loop):
society_simplex.change_generation()
society_blx.change_generation()
# 画像データ表示用処理
best_simplex = society_simplex.get_best_individual()
best_blx = society_blx.get_best_individual()
simplex_data.append(best_simplex.evaluate_value)
blx_data.append(best_blx.evaluate_value)
iterator_list.append(i)
print(i)
# 評価値の推移を表示
plot_evaluate_values(iterator_list, simplex_data, blx_data)
#!/usr/bin/env python3 # -*- coding: utf-8 -*- """ Created on Tue Apr 7 18:42:28 2020 @author: beca4397 """ from society import Society soc1 = Society()
from person import Person
from meme import Meme
from society import Society
SCREEN_SIZE = 1000
clock = pygame.time.Clock()
pygame.init()
pygame.display.set_caption('Memetic Evolution Simulation')
screen = pygame.display.set_mode((SCREEN_SIZE, SCREEN_SIZE), 0, 32)
missionary = Meme(spread=200)
cult = Meme(spread=5)
society_1 = Society(SCREEN_SIZE / 3, SCREEN_SIZE / 2)
society_2 = Society(2 * SCREEN_SIZE/3, SCREEN_SIZE/2)
societies = [society_1, society_2]
people = []
for _ in range(100):
people.append(Person(society=society_1))
for _ in range(10):
people.append(Person(society=society_1, meme=missionary))
for _ in range(100):
people.append(Person(society=society_2))
for _ in range(10):
people.append(Person(society=society_2, meme=cult))