def test_optimizer2():
optimizer = Optimizer()
run_id = 0
## set population size
optimizer.size = 40
optimizer.max_generations = 5
## if not given: outfile will default to using the strategy name
## optimizer.outfile ='retrace_optimized.xls'
## parameter space to search over
## momentum = entry momentum crossover
## average = moving average filter
## duration = trade holding period
param_list = [
dict(name='momentum', min_val=10, max_val=100, steps=32,
converter=int),
dict(name='average', min_val=20, max_val=200, steps=32, converter=int),
dict(name='duration', min_val=10, max_val=50, steps=16, converter=int)
]
optimizer.add_parameters(param_list)
while not optimizer.converged():
params_set = optimizer.generate_set()
for strategy_params in params_set:
log.debug('Testing: %s' % pprint.pformat(strategy_params))
sim_stats = test_retrace_strategy(strategy_params, run_id)
optimizer.score(sim_stats, run_id)
run_id += 1
optimizer.dump()
optimizer.write()
def test_optimizer3():
optimizer = Optimizer()
run_id = 0
## set population size
optimizer.size = 40
optimizer.max_generations = 5
optimizer.outfile = 'triple_optimized.xls'
## parameter space to search over
## mo1 = short term momentum crossover
## mo2 = medium term mo filter
## mo3 = long term mo filter
## duration = trade holding period
param_list = [
dict(name='mo1', min_val=5, max_val=50, steps=32, converter=int),
dict(name='mo2', min_val=60, max_val=100, steps=32, converter=int),
dict(name='mo3', min_val=110, max_val=200, steps=32, converter=int),
dict(name='duration', min_val=10, max_val=50, steps=16, converter=int)
]
optimizer.add_parameters(param_list)
while not optimizer.converged():
params_set = optimizer.generate_set()
for strategy_params in params_set:
log.debug('Testing: %s' % pprint.pformat(strategy_params))
sim_stats = test_triple_strategy(strategy_params, run_id)
optimizer.score(sim_stats, run_id)
run_id += 1
optimizer.dump()
optimizer.write()
def optimize():
optimizer = Optimizer()
optimizer.strategy_class = RSIStrategy
#optimizer.data_feed = DataFeedList(['20081210.SPY.1m.csv','20081211.SPY.1min.csv','20081212.SPY.1min.csv'],data_type='B')
#optimizer.data_feed = DataFeedList(['20081210.SPY.30s.csv','20081211.SPY.30s.csv','20081212.SPY.30s.csv'],data_type='I')
optimizer.data_feed = DataFeedList(['SPY.csv'], data_type='D')
## set population size
optimizer.size = 40
optimizer.max_generations = 50
optimizer.outfile = '%s_%s.xls' % (__file__[:-3],
datetime.now().strftime("%Y%m%d"))
optimizer.tolerance = 0.01
#optimizer.reset_on_EOD = False
## parameter space to search over
## rsi = rsi length
## top/btm = rsi buy sell thresholds
## average(optional) trend filter
## duration = trade duration
optimizer.add_parameter(
dict(name='rsi', min_val=10, max_val=40, steps=16, converter=int))
optimizer.add_parameter(
dict(name='top', min_val=60, max_val=80, steps=4, converter=int))
optimizer.add_parameter(
dict(name='btm', min_val=20, max_val=40, steps=4, converter=int))
optimizer.add_parameter(
dict(name='average', min_val=20, max_val=200, steps=64, converter=int))
optimizer.add_parameter(
dict(name='duration', min_val=5, max_val=20, steps=8, converter=int))
optimizer.run()
from RetraceStrategy import RetraceStrategy
from DataFeed import DataFeedDaily
from Optimizer import Optimizer
if __name__ == '__main__':
optimizer = Optimizer()
optimizer.strategy_class = RetraceStrategy
# optimizer.data_feed = DataFeedDaily('daily.SPY.csv')
optimizer.data_feed = DataFeedDaily('SPY.csv')
## set population size
optimizer.size = 40
optimizer.max_generations = 50
optimizer.outfile = 'optimize_retrace3.xls'
optimizer.reset_on_EOD = True
## parameter space to search over
## strategy_params for RetraceStrategy = dict(average,momentum,duration)
## momentum = entry momentum crossover
## average = moving average filter
## duration = trade holding period
param_list = [
dict(name='momentum', min_val=30, max_val=50, steps=16, converter=int),
dict(name='average', min_val=70, max_val=120, steps=32, converter=int),
dict(name='duration', min_val=10, max_val=20, steps=8, converter=int)
]
for p in param_list:
optimizer.add_parameter(p)
