def test_allocate_probabilities():
"""
Unit test for allocating cumulative probabilities.
"""
random.seed(42)
data = {}
for i in range(0, 100):
data[i] = {'id': i, 'population': random.randint(100, 1000)}
num_substations = [4]
iterations = 100
probabilities = [0.01]
all_results = simulation(data, num_substations, probabilities, iterations)
cp_scenarios = allocate_probabilities(all_results, num_substations,
probabilities)
assert len(cp_scenarios) == 1
assert cp_scenarios[0]['nodes'] == 4
assert cp_scenarios[0]['iteration'] == 96
assert cp_scenarios[0]['population'] == 3477
assert cp_scenarios[0]['substation_ids'] == [33, 26, 85, 91]
assert cp_scenarios[0]['rank'] == 1
assert cp_scenarios[0]['cum_probability'] == 0.01
data = {}
for i in range(0, 100):
data[i] = {'id': i, 'population': random.randint(100, 1000)}
num_substations = [7]
iterations = 100
probabilities = [0.5]
all_results = simulation(data, num_substations, probabilities, iterations)
cp_scenarios = allocate_probabilities(all_results, num_substations,
probabilities)
assert len(cp_scenarios) == 1
assert cp_scenarios[0]['nodes'] == 7
assert cp_scenarios[0]['iteration'] == 92
assert cp_scenarios[0]['population'] == 3551
assert cp_scenarios[0]['substation_ids'] == [89, 49, 63, 51, 31, 18, 83]
assert cp_scenarios[0]['rank'] == 50
assert cp_scenarios[0]['cum_probability'] == 0.495
num_substations = [4, 7, 14]
all_results = simulation(data, num_substations, probabilities, iterations)
cp_scenarios = allocate_probabilities(all_results, num_substations,
probabilities)
assert len(cp_scenarios) == 3
def test_simulation():
"""
Integration test for the main simulation generation function.
"""
random.seed(42)
data = {}
for i in range(0, 100):
data[i] = {'id': i, 'population': random.randint(100, 1000)}
num_substations = [4]
iterations = 100
probabilities = [0.5, 0.1, 0.01]
all_results = simulation(data, num_substations, probabilities, iterations)
assert len(all_results) == 100
max_iters = max([i['iteration'] for i in all_results])
assert max_iters == (iterations - 1
) #Remember: iterations run from 0 to 99.
max_nodes = max([len(i['substation_ids']) for i in all_results])
assert max_nodes == num_substations[0]
unique_nodes = list(set([i['nodes'] for i in all_results]))
assert unique_nodes == num_substations
def test_cascading_failures():
"""
Unit test for quantifying cascading failures.
"""
random.seed(42)
data = {}
for i in range(0, 100):
data[i] = {'id': i, 'population': random.randint(100, 1000)}
data_indirect = {}
for i in range(0, 100):
rand = random.randint(1, 2)
if rand == 1:
function = 'Railway Station'
elif rand == 2:
function = 'Gas Distribution or Storage'
else:
print('Did not recognize selected int')
data_indirect[i] = {
'dest_func': function,
}
num_substations = [4]
iterations = 1
probabilities = [0.01]
all_results = simulation(data, num_substations, probabilities, iterations)
all_results = cascading_failures(all_results, data_indirect)
assert all_results[0]['Railway Station'] == 2
assert all_results[0]['Gas Distribution or Storage'] == 2
return data_indirect
if __name__ == '__main__':
if not os.path.exists(RESULTS):
os.makedirs(RESULTS)
num_substations = [4, 7, 14]
iterations = 1000
probabilities = [0.5, 0.1, 0.01]
data = load_data()
all_results = simulation(data, num_substations, probabilities, iterations)
data_indirect = load_indirect_data()
all_results = cascading_failures(all_results, data_indirect)
cp_scenarios = allocate_probabilities(all_results, num_substations,
probabilities)
all_results = pd.DataFrame(all_results)
path = os.path.join(RESULTS, 'all_results.csv')
all_results.to_csv(path, index=False)
cp_scenarios = pd.DataFrame(cp_scenarios)
path = os.path.join(RESULTS, 'cp_scenarios.csv')
cp_scenarios.to_csv(path, index=False)