def constructSimulation(graph, num_cars, link_len_range, run_graphics,
debug_construct, debug_sim):
"""Builds and runs the simulation given the input parameters
graph: - a correct symbolic representation that is being simulated
num_cars: - number of cars to be in the simulation
link_len_range: - the range of values that the length of the roads can take
run_graphics: - boolean indicating whether or not the graphical overlay should be generated
debug_construct: - boolean indicating whether construction is being debugged
debug_sim: - boolean indicating whether simulator is being debugged"""
num_junctions = len(graph.node_list)
num_roads = len(graph.link_list)
road_angles = [x.angle for x in graph.link_list]
road_lengths = setLinkLengths(graph.link_list, link_len_range,
debug_construct)
car_goals = [random.randint(0, num_junctions) for x in range(num_roads)]
junc_pairs = []
pre, post = None, None
for link in graph.link_list:
pre = link.from_node
post = link.to_node
if link.angle > 90 and link.angle <= 270:
junc_pairs.append((post.label, pre.label))
link.angle -= 180
else:
junc_pairs.append((pre.label, post.label))
print("{}-{}\t{}\t{}".format(link.from_node.label, link.to_node.label,
junc_pairs[-1], link.length))
if not debug_construct:
simulator.runSimulation(num_junctions,num_roads,num_cars,road_angles,road_lengths,\
junc_pairs,car_goals,run_graphics,debug_sim)
def main():
welcome()
coins = fetchCoins()
currency, quantity = inputBuy()
try:
price = coins[currency]['price']
except Exception as e:
dramaticTyping("Invalid currency entered, please try again \n")
inputBuy()
runSimulation(coins[currency]['price'], quantity, currency)
quitMenu()
def test_largePortfolio(self):
i = 1000000
result = runSimulation(30, i, i * 0.04,
((ConstantWithdrawalAmountStrategy(i * .04),
Assets(.5, .5), 1.0), ), 1926, 1997)
self.assertAlmostEqual(result.getSuccessRate(), .95, delta=.005)
def test_constantPercentRegression(self):
result = runSimulation(
30, 1.0, .1,
((ConstantPercentWithdrawalStrategy(.1), Assets(.5, .5), 1.0), ),
1926, 1997)
self.assertLessEqual(result.getSuccessRate(), .95)
def test_ramp(self):
i = 1000000
result = runSimulation(30, i, i * 0.04,
((ConstantWithdrawalAmountStrategy(i * .04),
LinearRamp(Assets(.8, .2), Assets(.2, .8),
(30 + 1) * 12), 1.0), ), 1926,
1997)
self.assertAlmostEqual(result.getSuccessRate(), .98, delta=.005)
def trinity(length,
initialWithdrawal,
initialPortfolio,
ignoreInflation=False,
maxYear=trinityMaxYear):
return runSimulation(length, 1000000, initialWithdrawal,
((ConstantWithdrawalAmountStrategy(initialWithdrawal),
initialPortfolio.allocation, 1.0), ), 1926, maxYear,
ignoreInflation)
def test_multipleStrategies(self):
i = 1000000
result = runSimulation(30, i, i * 0.04,
((ConstantWithdrawalAmountStrategy(
i * .04 * .6), Assets(.5, .5), .6),
(ConstantWithdrawalAmountStrategy(
i * .04 * .4), Assets(.5, .5), .4)), 1926,
1997)
self.assertAlmostEqual(result.getSuccessRate(), .95, delta=.005)
def test_multipleDifferentStrategies(self):
i = 1000000
result = runSimulation(
30, i, 0.0,
((ConstantWithdrawalAmountStrategy(i * .04 * .5), Assets(.5,
.5), .5),
(ConstantPercentWithdrawalStrategy(.10), Assets(1.0, 0.0), .5)),
1926, 2009)
# it always works because you never run out with a percent withdrwawal strategy
self.assertAlmostEqual(result.getSuccessRate(), 1.0, delta=.005)
def test_basicWithdrawal(self):
retirementLength = 30
initialPortfolio = 1 * 1000 * 1000
result = runSimulation(
retirementLength, initialPortfolio, .05 * initialPortfolio * .5,
((Vpw(.04, retirementLength, .75), Assets(.5, .5), 1.0), ), 1926,
2009)
result.getSimResults()
print(result)
def test_noOptions(self):
retirementLength = 30
result = runSimulation(
retirementLength, 1 * 1000 * 1000, 20000,
((GuytonKlinger(40000, False, False, False, False,
retirementLength), Assets(.5, .5), 1.0), ), 1926,
2009)
result.getSimResults()
print(result)
def update_data(self):
"""Called each time that any watched property changes.
We grab the parameters from the form and run the
simulation in a 'pure' function
"""
N = 120
# Get the current slider values
p = self.prepay_probability.value
d = self.default_probability.value
self.source.data = runSimulation(N,p,d)
from assets import Assets
from simulator import runSimulation
from strategies.guyton_klinger import GuytonKlinger
"""
This is just a playground file. Useful for running the profiler, first attempt at running the code, etc.
"""
retirementLength = 30
initialPortfolio = 1 * 1000 * 1000
result = runSimulation(
retirementLength, initialPortfolio, .05 * initialPortfolio * .5,
((GuytonKlinger(.05 * initialPortfolio, True, True, True, True,
retirementLength), Assets(.5, .5), 1.0), ), 1926, 2009)
result.getSimResults()
print(result)
from simulator import runSimulation, runMany N = 120 p = 0.01 d = 0.01 dic = runSimulation(N,p,d) str1 = "using a prepay rate of " + str(p) str2 = " and a default rate of " + str(d) print str1 + str2 print dic['x'] print dic['y'] runMany(120)
def test_notConjuringMoney(self):
assets = Assets(.5, .5)
i = 1000000
simLength = 30
strategies = [
((GuytonKlinger(i * .04, True, True, True, True,
simLength), assets, 1.0), True),
((ConstantWithdrawalAmountStrategy(i * .04), assets, 1.0), True),
((ConstantPercentWithdrawalStrategy(.04), assets, 1.0), True),
((Vpw(.02, simLength, .04), assets, 1.0), True),
((HebelerAuto(55), assets, 1.0), True),
((LeprechaunGoldStrategy(.04), assets, 1.0), False),
((YearlyLeprechaunGoldStrategy(.04), assets, 1.0), False)
]
class mblp: # stands for "Make better lambdas python"
prePortfolio = Portfolio(assets, i)
growth = None
def pre(observedPrePortfolioValue, observedGrowth, strategies):
if len(strategies) != 1:
raise RuntimeError
mblp.growth = max(observedGrowth)
if hasattr(strategies[0], 'yearGrowthAccumulator'):
mblp.growth = max(mblp.growth,
max(strategies[0].yearGrowthAccumulator),
1.0)
mblp.prePortfolio.value = observedPrePortfolioValue
def post(observedPortfolioValue, observedWithdrawal):
mblp.prePortfolio.value -= observedWithdrawal
mblp.prePortfolio.value = mblp.prePortfolio.value * mblp.growth
maxPossible = mblp.prePortfolio.value
observed = observedPortfolioValue
if mblp.expectSuccess:
self.assertGreaterEqual(maxPossible + .00000001,
observed,
msg="Leprechauns detected.")
else:
if maxPossible + .00000001 < observed:
mblp.sawAFailure = True
for s in strategies:
mblp.prePortfolio = Portfolio(assets, i)
mblp.growth = None
mblp.sawAFailure = False
mblp.expectSuccess = s[1]
testhook = {"pre": pre, "post": post}
runSimulation(simLength,
i,
i * 0.02, (s[0], ),
1926,
2009,
ignoreInfation=True,
testCallback=testhook)
if not mblp.expectSuccess and not mblp.sawAFailure:
self.fail("Goblins detected.")
def simulations():
minYear = int(request.json.get("min_year", 1926))
maxYear = int(request.json.get("max_year", 2009))
failureThreshold = float(request.json.get("failure_threshold", 20 * 1000))
initialPortfolioValue = float(
request.json.get("initial_portfolio_value", 1 * 1000 * 1000))
retirementLength = int(request.json.get("retirement_length", 30))
strategies = []
for s in request.json["strategies"]:
args = s["args"]
if type(s["asset_allocation"]) is dict:
if s["asset_allocation"]["type"].lower() == "linear_ramp":
allocation = LinearRamp(
Assets([float(f) for f in s["asset_allocation"]["start"]]),
Assets([float(f) for f in s["asset_allocation"]["end"]]),
(retirementLength + 1) * 12)
else:
raise NotImplemented("Unrecognized ramp: {0}".format(
s["asset_allocation"]["type"].lower()))
else:
allocation = Assets([float(f) for f in s["asset_allocation"]])
weight = float(s.get("weight", 1.0))
strategyType = s["type"]
if strategyType.lower() == "guyton_klinger":
strategy = GuytonKlinger(float(args.get("initial_amount")),
bool(args.get("gk_pmr")),
bool(args.get("gk_wr")),
bool(args.get("gk_cpr")),
bool(args.get("gk_pr")), retirementLength)
elif strategyType.lower() == "const_amount":
strategy = ConstantWithdrawalAmountStrategy(float(args["amount"]))
elif strategyType.lower() == "const_percent":
strategy = ConstantPercentWithdrawalStrategy(float(
args["percent"]))
elif strategyType.lower() == "hebeler_autopilot":
strategy = HebelerAuto(int(args["age"]))
elif strategyType.lower() == "vpw":
strategy = Vpw(float(args["expected_return_percent"]),
int(retirementLength),
float(args["drawdown_percent"]))
else:
raise NotImplemented(
"Unrecognized strategy: {0}".format(strategyType))
strategies.append((strategy, allocation, weight))
result = runSimulation(retirementLength, initialPortfolioValue,
failureThreshold, strategies, minYear, maxYear)
results = result.getSimResults()
initialWithdrawal = results[0]["withdrawals"][0]
if strategyType.lower() == "const_percent":
initialWithdrawal = initialPortfolioValue * float(args["percent"])
if strategyType.lower() == "guyton_klinger":
initialWithdrawal = float(args.get("initial_amount"))
return flask.jsonify(initial_withdrawal_amt=initialWithdrawal,
simulation_start=minYear,
simulation_end=maxYear,
results=results,
stats=result.getStats(),
dist_stats=result.getDistStats(),
yearly_stats=result.getYearlyStats())
