def TaskFour():
problem_bounds = (-50,150)
max_evals = 10000
desired_goal_value = (0.1)**6
divergence_at = 1000
iterations = 11
#population_values = [30, 50, 100, 200] #[30, 50, 100, 200]
mutation_values = [0.01, 0.03, 0.06, 0.1] #[0.1, 0.3, 0.6, 0.9]
#for pop_value in population_values:
for mut_value in mutation_values:
results = []
for i in range(iterations):
GA = GeneticAlgorithm(goal_function=f6, dimensions=2, problem_bounds=problem_bounds,
fitness_bounds=(0,100), population_size=100, display="binary", precision=4, p_of_mutation=mut_value,
max_evaluations=max_evals, reach_goal=desired_goal_value, no_improvement_limit=divergence_at)
point, goal_value = GA.SolveProblem()
results.append((goal_value, point))
print goal_value
results = SortGoalAndPoint(results)
count = CountSuccesses(results, desired_goal_value)
PrintResults(results, count, iterations)
def ParamOpti():
problem_bounds = (-50,150)
max_evals = 10000
desired_goal_value = (0.1)**6
iterations = 21
population_values = [30, 50, 100, 200] #[30, 50, 100, 200]
mutation_values = [0.01,0.05] #[0.01,0.05,0.1,0.2,0.5]
precision_values = [0,2,4,8,16]
for opti_value in mutation_values:
results = []
for i in range(iterations):
GA = GeneticAlgorithm(goal_function=f6, dimensions=2, problem_bounds=problem_bounds,
fitness_bounds=(0,100), population_size=50, display="binary", precision=8, p_of_mutation=opti_value,
max_evaluations=max_evals, reach_goal=desired_goal_value)
point, goal_value = GA.SolveProblem()
results.append((goal_value, point))
results = SortGoalAndPoint(results)
count = CountSuccesses(results, desired_goal_value)
PrintResults(results, count, iterations)
print "optimization_value:",
print opti_value
def TaskOne():
problem_bounds = (-50,150)
max_evals = 10000
desired_goal_value = (0.1)**6
divergence_at = 1000
iterations = 11
results = []
for i in range(iterations):
# GA = GeneticAlgorithm(goal_function=f1, dimensions=2, problem_bounds=problem_bounds,
# fitness_bounds=(0,100), population_size=100, display="float", precision=8, p_of_mutation=0.01,
# max_evaluations=max_evals, reach_goal=desired_goal_value, no_improvement_limit=divergence_at)
# GA = GeneticAlgorithm(goal_function=f3, dimensions=5, problem_bounds=problem_bounds,
# fitness_bounds=(0,100), population_size=4, display="float", precision=8, p_of_mutation=0.01,
# max_evaluations=max_evals, reach_goal=desired_goal_value, no_improvement_limit=divergence_at)
# GA = GeneticAlgorithm(goal_function=f6, dimensions=2, problem_bounds=problem_bounds,
# fitness_bounds=(0,100), population_size=100, display="binary", precision=4, p_of_mutation=0.01,
# max_evaluations=max_evals, reach_goal=desired_goal_value, no_improvement_limit=divergence_at)
GA = GeneticAlgorithm(goal_function=f7, dimensions=2, problem_bounds=problem_bounds,
fitness_bounds=(0,100), population_size=100, display="float", precision=0, p_of_mutation=0.01,
max_evaluations=max_evals, reach_goal=desired_goal_value, no_improvement_limit=divergence_at)
point, goal_value = GA.SolveProblem()
results.append((goal_value, point))
results = SortGoalAndPoint(results)
count = CountSuccesses(results, desired_goal_value)
PrintResults(results, count, iterations)
def test_GeneticAlgorithm_Binary_MaxEval_f3():
algo = GeneticAlgorithm(goal_function=f3,
dimensions=2,
problem_bounds=(-50, 150),
fitness_bounds=(0, 100),
population_size=100,
display="binary",
precision=0,
p_of_mutation=0.01,
max_evaluations=1000)
result = algo.SolveProblem()
assert True #the result cannot be interpreted
def test_GeneticAlgorithm_Float_ConvergedOnAPoint_f3():
algo = GeneticAlgorithm(goal_function=f3,
dimensions=2,
problem_bounds=(-50, 150),
fitness_bounds=(0, 100),
population_size=100,
display="float",
precision=0,
p_of_mutation=0.01,
no_improvement_limit=50)
result = algo.SolveProblem()
assert True #the result cannot be interpreted
def test_GeneticAlgorithm_Float_GoalValue_f3():
algo = GeneticAlgorithm(goal_function=f3,
dimensions=2,
problem_bounds=(-50, 150),
fitness_bounds=(0, 100),
population_size=100,
display="float",
precision=0,
p_of_mutation=0.01,
reach_goal=(0.1)**3)
result = algo.SolveProblem()
assert True #the result cannot be interpreted
def SIAMHundredDigitChallenge():
problem_bounds = (0,20)
max_evals = 100000
desired_goal_value = (-3.3068686473)
divergence_at = 5000
iterations = 501
results = []
for i in range(iterations):
GA = GeneticAlgorithm(goal_function=SIAMf, dimensions=2, problem_bounds=problem_bounds,
fitness_bounds=(0,100), population_size=200, display="float", precision=0, p_of_mutation=0.01,
max_evaluations=max_evals, reach_goal=desired_goal_value, no_improvement_limit=divergence_at)
point, goal_value = GA.SolveProblem()
results.append((goal_value, point))
results = SortGoalAndPoint(results)
count = CountSuccesses(results, desired_goal_value)
PrintResults(results, count, iterations)
def TaskFive():
"""
The tournament size is set at k=3, but it can be increased for testing purposes within the algorithm.
WARNING: always return the tournmanet size after testing.
"""
problem_bounds = (-50,150)
max_evals = 10000
desired_goal_value = (0.1)**6
divergence_at = 1000
iterations = 21
results = []
for i in range(iterations):
GA = GeneticAlgorithm(goal_function=f6, dimensions=2, problem_bounds=problem_bounds,
fitness_bounds=(0,100), population_size=100, display="binary", precision=0, p_of_mutation=0.01,
max_evaluations=max_evals, reach_goal=desired_goal_value, no_improvement_limit=divergence_at)
point, goal_value = GA.SolveProblem()
results.append((goal_value, point))
print goal_value
results = SortGoalAndPoint(results)
count = CountSuccesses(results, desired_goal_value)
PrintResults(results, count, iterations)