def test_simulated_annealing_for_objective_function_starting_temp(
slots, events):
def objective_function(array):
return of.efficiency_capacity_demand_difference(slots, events, array)
array = np.array([
[1, 0, 0, 0, 0, 0, 0],
[0, 0, 0, 0, 1, 0, 0],
[0, 1, 0, 0, 0, 0, 0]
])
assert objective_function(array) == 400
np.random.seed(0)
X = simulated_annealing(initial_array=array,
objective_function=objective_function,
initial_temperature=0,
max_iterations=10)
assert objective_function(X) == 250
np.random.seed(0)
X = simulated_annealing(initial_array=array,
objective_function=objective_function,
initial_temperature=1000,
max_iterations=10)
assert objective_function(X) == 250
def test_simulated_annealing_for_valid_solution_warning_raised(slots, events):
"""
Test that a warning is given if a lower bound is passed and not reached in
given number of iterations.
"""
def objective_function(array):
return len(list(array_violations(array, events, slots)))
array = np.array([
[1, 0, 0, 0, 0, 0, 0],
[0, 0, 1, 0, 0, 0, 0],
[0, 0, 0, 0, 0, 1, 0]
])
assert objective_function(array) == 2
np.random.seed(0)
with warnings.catch_warnings(record=True) as w:
X = simulated_annealing(initial_array=array,
objective_function=objective_function,
lower_bound=0,
max_iterations=1)
assert objective_function(X) == 1
assert len(w) == 1
def test_simulated_annealing_for_objective_function_cooldown_rate(
slots, events):
def objective_function(array):
return of.efficiency_capacity_demand_difference(slots, events, array)
array = np.array([
[1, 0, 0, 0, 0, 0, 0],
[0, 0, 0, 0, 1, 0, 0],
[0, 1, 0, 0, 0, 0, 0]
])
np.random.seed(0)
X = simulated_annealing(initial_array=array,
objective_function=objective_function,
cooldown_rate=.1,
max_iterations=10)
assert objective_function(X) == 250
def test_simulated_annealing_for_valid_solution(slots, events):
def objective_function(array):
return len(list(array_violations(array, events, slots)))
array = np.array([
[1, 0, 0, 0, 0, 0, 0],
[0, 0, 1, 0, 0, 0, 0],
[0, 0, 0, 0, 0, 1, 0]
])
assert objective_function(array) == 2
np.random.seed(0)
X = simulated_annealing(initial_array=array,
lower_bound=0,
objective_function=objective_function,
max_iterations=150)
assert objective_function(X) == 0
def test_simulated_annealing_for_obj_function_with_criteria(slots, events):
def objective_function(array):
return of.efficiency_capacity_demand_difference(slots, events, array)
def acceptance_criteria(array):
return sum(array[:, 3])
array = np.array([
[1, 0, 0, 0, 0, 0, 0],
[0, 0, 0, 0, 1, 0, 0],
[0, 1, 0, 0, 0, 0, 0]
])
assert acceptance_criteria(array) == 0
assert objective_function(array) == 400
np.random.seed(0)
X = simulated_annealing(initial_array=array,
objective_function=objective_function,
acceptance_criteria=acceptance_criteria,
max_iterations=100)
assert objective_function(X) == 100