def turtle( asset, start_date, end_date, systems, config):
rollrule = config['rollrule']
nearby = config['nearby']
file_prefix = config['file_prefix'] + '_' + asset + '_'
start_d = misc.day_shift(start_date, '-'+str(max([ max(sys) for sys in systems]))+'b')
ddf = misc.nearby(asset, nearby, start_d, end_date, rollrule, 'd', need_shift=True)
mdf = misc.nearby(asset, nearby, start_date, end_date, rollrule, 'm', need_shift=True)
#ddf = dh.conv_ohlc_freq(mdf, 'D')
output = {}
for ix, sys in enumerate(systems):
config['signals'] = sys[:3]
config['max_loss'] = sys[3]
config['max_pos'] = sys[4]
(res, closed_trades, ts) = turtle_sim( ddf, mdf, config)
output[ix] = res
print 'saving results for scen = %s' % str(ix)
all_trades = {}
for i, tradepos in enumerate(closed_trades):
all_trades[i] = strat.tradepos2dict(tradepos)
fname = file_prefix + str(ix) + '_trades.csv'
trades = pd.DataFrame.from_dict(all_trades).T
trades.to_csv(fname)
fname = file_prefix + str(ix) + '_dailydata.csv'
ts.to_csv(fname)
fname = file_prefix + 'stats.csv'
res = pd.DataFrame.from_dict(output)
res.to_csv(fname)
return
def dual_thrust( asset, start_date, end_date, scenarios, config):
nearby = config['nearby']
rollrule = config['rollrule']
start_d = misc.day_shift(start_date, '-2b')
file_prefix = config['file_prefix'] + '_' + asset + '_'
ddf = misc.nearby(asset, nearby, start_d, end_date, rollrule, 'd', need_shift=True)
mdf = misc.nearby(asset, nearby, start_d, end_date, rollrule, 'm', need_shift=True)
#ddf = dh.conv_ohlc_freq(mdf, 'D')
output = {}
for ix, s in enumerate(scenarios):
config['win'] = s[1]
config['k'] = s[0]
config['m'] = s[2]
(res, closed_trades, ts) = dual_thrust_sim( ddf, mdf, config)
output[ix] = res
print 'saving results for scen = %s' % str(ix)
all_trades = {}
for i, tradepos in enumerate(closed_trades):
all_trades[i] = strat.tradepos2dict(tradepos)
fname = file_prefix + str(ix) + '_trades.csv'
trades = pd.DataFrame.from_dict(all_trades).T
trades.to_csv(fname)
fname = file_prefix + str(ix) + '_dailydata.csv'
ts.to_csv(fname)
fname = file_prefix + 'stats.csv'
res = pd.DataFrame.from_dict(output)
res.to_csv(fname)
return
def dual_thrust( asset, start_date, end_date, scenarios, config):
nearby = config['nearby']
rollrule = config['rollrule']
start_d = misc.day_shift(start_date, '-4b')
file_prefix = config['file_prefix'] + '_' + asset + '_'
ddf = misc.nearby(asset, nearby, start_d, end_date, rollrule, 'd', need_shift=True)
mdf = misc.nearby(asset, nearby, start_d, end_date, rollrule, 'm', need_shift=True)
mdf = backtest.cleanup_mindata(mdf, asset)
#ddf = dh.conv_ohlc_freq(mdf, 'D')
output = {}
for ix, s in enumerate(scenarios):
config['win'] = s[1]
config['k'] = s[0]
config['m'] = s[2]
(res, closed_trades, ts) = dual_thrust_sim( ddf, mdf, config)
output[ix] = res
print 'saving results for scen = %s' % str(ix)
all_trades = {}
for i, tradepos in enumerate(closed_trades):
all_trades[i] = strat.tradepos2dict(tradepos)
fname = file_prefix + str(ix) + '_trades.csv'
trades = pd.DataFrame.from_dict(all_trades).T
trades.to_csv(fname)
fname = file_prefix + str(ix) + '_dailydata.csv'
ts.to_csv(fname)
fname = file_prefix + 'stats.csv'
res = pd.DataFrame.from_dict(output)
res.to_csv(fname)
return
def r_breaker( asset, start_date, end_date, scenarios, freqs, config):
nearby = config['nearby']
rollrule = config['rollrule']
start_d = misc.day_shift(start_date, '-1b')
file_prefix = config['file_prefix'] + '_' + asset + '_'
ddf = misc.nearby(asset, nearby, start_date, end_date, rollrule, 'd', need_shift=True)
mdf = misc.nearby(asset, nearby, start_date, end_date, rollrule, 'm', need_shift=True)
mdf = backtest.cleanup_mindata(mdf, asset)
#ddf = dh.conv_ohlc_freq(mdf, 'D')
output = {}
for ix, freq in enumerate(freqs):
if freq !='1min':
df = dh.conv_ohlc_freq(mdf, freq)
else:
df = mdf
for iy, k in enumerate(scenarios):
idx = ix*10+iy
config['k'] = k
(res, closed_trades, ts) = r_breaker_sim( ddf, df, config)
output[idx] = res
print 'saving results for scen = %s' % str(idx)
all_trades = {}
for i, tradepos in enumerate(closed_trades):
all_trades[i] = strat.tradepos2dict(tradepos)
fname = file_prefix + str(idx) + '_trades.csv'
trades = pd.DataFrame.from_dict(all_trades).T
trades.to_csv(fname)
fname = file_prefix + str(idx) + '_dailydata.csv'
ts.to_csv(fname)
fname = file_prefix + 'stats.csv'
res = pd.DataFrame.from_dict(output)
res.to_csv(fname)
return
def chanbreak( asset, start_date, end_date, freqs, windows, config):
nearby = config['nearby']
rollrule = config['rollrule']
file_prefix = config['file_prefix'] + '_' + asset + '_'
ddf = misc.nearby(asset, nearby, start_d, end_date, rollrule, 'd', need_shift=True)
mdf = misc.nearby(asset, nearby, start_d, end_date, rollrule, 'm', need_shift=True)
mdf = backtest.cleanup_mindata(mdf, asset)
output = {}
for ix, freq in enumerate(freqs):
config['freq'] = freq
for iy, win in enumerate(windows):
idx = ix*10+iy
config['win'] = win
(res, closed_trades, ts) = chanbreak_sim( mdf, ddf, config)
output[idx] = res
print 'saving results for scen = %s' % str(idx)
all_trades = {}
for i, tradepos in enumerate(closed_trades):
all_trades[i] = strat.tradepos2dict(tradepos)
fname = file_prefix + str(idx) + '_trades.csv'
trades = pd.DataFrame.from_dict(all_trades).T
trades.to_csv(fname)
fname = file_prefix + str(idx) + '_dailydata.csv'
ts.to_csv(fname)
fname = file_prefix + 'stats.csv'
res = pd.DataFrame.from_dict(output)
res.to_csv(fname)
return
def dual_thrust(asset, start_date, end_date, scenarios, config):
nearby = config["nearby"]
rollrule = config["rollrule"]
start_d = misc.day_shift(start_date, "-4b")
file_prefix = config["file_prefix"] + "_" + asset + "_"
ddf = misc.nearby(asset, nearby, start_d, end_date, rollrule, "d", need_shift=True)
mdf = misc.nearby(asset, nearby, start_d, end_date, rollrule, "m", need_shift=True)
mdf = backtest.cleanup_mindata(mdf, asset)
# ddf = dh.conv_ohlc_freq(mdf, 'D')
output = {}
for ix, s in enumerate(scenarios):
config["win"] = s[1]
config["k"] = s[0]
config["m"] = s[2]
(res, closed_trades, ts) = dual_thrust_sim(ddf, mdf, config)
output[ix] = res
print "saving results for scen = %s" % str(ix)
all_trades = {}
for i, tradepos in enumerate(closed_trades):
all_trades[i] = strat.tradepos2dict(tradepos)
fname = file_prefix + str(ix) + "_trades.csv"
trades = pd.DataFrame.from_dict(all_trades).T
trades.to_csv(fname)
fname = file_prefix + str(ix) + "_dailydata.csv"
ts.to_csv(fname)
fname = file_prefix + "stats.csv"
res = pd.DataFrame.from_dict(output)
res.to_csv(fname)
return
def load_data(self, idx):
asset = self.sim_assets[idx]
for prod in asset:
mdf = misc.nearby(prod, self.config['nearby'], self.config['start_date'], self.config['end_date'],
self.config['rollrule'], 'm', need_shift = self.need_shift, database=self.config.get('database', 'hist_data'))
mdf = misc.cleanup_mindata(mdf, prod)
self.min_data[prod] = mdf
def fisher_swing(asset, start_date, end_date, freqs, windows, config):
nearby = config["nearby"]
rollrule = config["rollrule"]
file_prefix = config["file_prefix"] + "_" + asset + "_"
df = misc.nearby(asset, nearby, start_date, end_date, rollrule, "m", need_shift=True)
df = backtest.cleanup_mindata(df, asset)
output = {}
for ix, freq in enumerate(freqs):
xdf = dh.conv_ohlc_freq(df, freq)
for iy, win in enumerate(windows):
idx = ix * 10 + iy
config["win"] = win
config["freq"] = freq
(res, closed_trades, ts) = fisher_swing_sim(df, xdf, config)
output[idx] = res
print "saving results for scen = %s" % str(idx)
all_trades = {}
for i, tradepos in enumerate(closed_trades):
all_trades[i] = strat.tradepos2dict(tradepos)
fname = file_prefix + str(idx) + "_trades.csv"
trades = pd.DataFrame.from_dict(all_trades).T
trades.to_csv(fname)
fname = file_prefix + str(idx) + "_dailydata.csv"
ts.to_csv(fname)
fname = file_prefix + "stats.csv"
res = pd.DataFrame.from_dict(output)
res.to_csv(fname)
return
def fisher_swing(asset, start_date, end_date, freqs, windows, config):
nearby = config['nearby']
rollrule = config['rollrule']
file_prefix = config['file_prefix'] + '_' + asset + '_'
df = misc.nearby(asset,
nearby,
start_date,
end_date,
rollrule,
'm',
need_shift=True)
df = backtest.cleanup_mindata(df, asset)
output = {}
for ix, freq in enumerate(freqs):
xdf = dh.conv_ohlc_freq(df, freq)
for iy, win in enumerate(windows):
idx = ix * 10 + iy
config['win'] = win
config['freq'] = freq
(res, closed_trades, ts) = fisher_swing_sim(df, xdf, config)
output[idx] = res
print 'saving results for scen = %s' % str(idx)
all_trades = {}
for i, tradepos in enumerate(closed_trades):
all_trades[i] = strat.tradepos2dict(tradepos)
fname = file_prefix + str(idx) + '_trades.csv'
trades = pd.DataFrame.from_dict(all_trades).T
trades.to_csv(fname)
fname = file_prefix + str(idx) + '_dailydata.csv'
ts.to_csv(fname)
fname = file_prefix + 'stats.csv'
res = pd.DataFrame.from_dict(output)
res.to_csv(fname)
return
def get_cont_data(asset,
start_date,
end_date,
freq='1m',
nearby=1,
rollrule='-10b'):
if nearby == 0:
mdf = mysqlaccess.load_min_data_to_df('fut_min',
asset,
start_date,
end_date,
minid_start=300,
minid_end=2114,
database='hist_data')
mdf['contract'] = asset
else:
mdf = misc.nearby(asset,
nearby,
start_date,
end_date,
rollrule,
'm',
need_shift=True,
database='hist_data')
mdf = backtest.cleanup_mindata(mdf, asset)
xdf = dh.conv_ohlc_freq(mdf,
freq,
extra_cols=['contract'],
bar_func=dh.bar_conv_func2)
return xdf
def load_data(self, idx):
asset = self.sim_assets[idx]
for prod in asset:
if prod in self.data_store:
continue
ticker = prod
if '$' not in ticker:
ticker_sp = [ticker, 'spot']
else:
ticker_sp = ticker.split('$')
ticker = ticker_sp[0]
postfix = '_daily'
if self.sim_freq == 'm':
postfix = '_min'
dbtable = ticker_sp[-1] + postfix
if ticker_sp[-1] in ['spot']:
field_id = 'spotID'
elif ticker_sp[-1] in ['ccy']:
field_id = 'instID'
if len(ticker_sp) > 2:
nb = int(ticker_sp[1])
if len(ticker_sp) > 3:
rollrule = ticker_sp[2]
else:
rollrule = '-1b'
df = misc.nearby(ticker,
nb,
self.config['start_date'],
self.config['end_date'],
rollrule,
self.sim_freq,
need_shift=self.need_shift,
database=self.config.get(
'dbconfig',
dbaccess.dbconfig)['database'])
else:
cnx = dbaccess.connect(
**self.config.get('dbconfig', dbaccess.dbconfig))
if self.sim_freq == 'd':
df = dbaccess.load_daily_data_to_df(
cnx,
dbtable,
ticker,
self.config['start_date'],
self.config['end_date'],
index_col='date',
field=field_id)
else:
minid_start = 1500
minid_end = 2114
if ticker in misc.night_session_markets:
minid_start = 300
df = dbaccess.load_min_data_to_df(
cnx, dbtable, ticker, self.config['start_date'],
self.config['end_date'], minid_start, minid_end)
df['contract'] = ticker
if self.sim_freq == 'm':
df = misc.cleanup_mindata(df, ticker)
df.columns = [(prod, col) for col in df.columns]
self.data_store[prod] = df
def get_data(spotID,
start,
end,
spot_table='spot_daily',
name=None,
index_col='date',
fx_pair=None,
field='spotID',
args=None):
cnx = dbaccess.connect(**dbaccess.dbconfig)
if args:
args['start_date'] = start
args['end_date'] = end
df = misc.nearby(spotID, **args)
else:
df = dbaccess.load_daily_data_to_df(cnx,
spot_table,
spotID,
start,
end,
index_col=None,
field=field)
if isinstance(df[index_col][0], basestring):
if len(df[index_col][0]) > 12:
df[index_col] = df[index_col].apply(
lambda x: datetime.datetime.strptime(x, "%Y-%m-%d %H:%M:%S"
).date())
else:
df[index_col] = df[index_col].apply(
lambda x: datetime.datetime.strptime(x, "%Y-%m-%d").date())
df = df.set_index(index_col)
if name:
col_name = name
else:
col_name = spotID
if field == 'ccy':
df = df[df.tenor == '0W']
data_field = 'rate'
elif field == 'spotID':
data_field = 'close'
df = df[[data_field]]
df.rename(columns={data_field: col_name}, inplace=True)
if fx_pair:
fx = dbaccess.load_daily_data_to_df(cnx,
'fx_daily',
fx_pair,
start,
end,
index_col=None,
field='ccy')
fx = fx[fx['tenor'] == '0W']
if isinstance(fx[index_col][0], basestring):
fx[index_col] = fx[index_col].apply(
lambda x: datetime.datetime.strptime(x, "%Y-%m-%d %H:%M:%S"
).date())
fx = fx.set_index(index_col)
fx = fx.set_index(index_col)
df[col_name] = df[col_name] / fx['rate']
return df
def load_data(self, idx):
asset = self.sim_assets[idx]
for prod in asset:
df = misc.nearby(prod, self.config['nearby'], self.config['start_date'], self.config['end_date'],
self.config['rollrule'], self.sim_freq, need_shift = self.need_shift)
if self.sim_freq == 'm':
df = misc.cleanup_mindata(df, prod)
self.data_store[prod] = df
def get_cont_data(asset, start_date, end_date, freq = '1m', nearby = 1, rollrule = '-10b'):
cnx = dbaccess.connect(**dbaccess.hist_dbconfig)
if nearby == 0:
mdf = dbaccess.load_min_data_to_df(cnx, 'fut_min', asset, start_date, end_date, minid_start = 300, minid_end = 2114, database = 'hist_data')
mdf['contract'] = asset
else:
mdf = misc.nearby(asset, nearby, start_date, end_date, rollrule, 'm', need_shift=True, database = 'hist_data')
mdf = misc.cleanup_mindata(mdf, asset)
xdf = dh.conv_ohlc_freq(mdf, freq, extra_cols = ['contract'], bar_func = dh.bar_conv_func2)
return xdf
def load_data(self, idx):
asset = self.sim_assets[idx]
for prod in asset:
df = misc.nearby(prod,
self.config['nearby'],
self.config['start_date'],
self.config['end_date'],
self.config['rollrule'],
self.sim_freq,
need_shift=self.need_shift)
if self.sim_freq == 'm':
df = misc.cleanup_mindata(df, prod)
self.data_store[prod] = df
def load_data(self, idx):
asset = self.sim_assets[idx]
for prod in asset:
if prod in self.data_store:
continue
ticker = prod
if '$' not in ticker:
ticker_sp = [ticker, 'spot']
else:
ticker_sp = ticker.split('$')
ticker = ticker_sp[0]
postfix = '_daily'
if self.sim_freq == 'm':
postfix = '_min'
dbtable = ticker_sp[-1] + postfix
if ticker_sp[-1] in ['spot']:
field_id = 'spotID'
elif ticker_sp[-1] in ['ccy']:
field_id = 'instID'
if len(ticker_sp) > 2:
nb = int(ticker_sp[1])
if len(ticker_sp) > 3:
rollrule = ticker_sp[2]
else:
rollrule = '-1b'
df = misc.nearby(ticker, nb, self.config['start_date'], self.config['end_date'], rollrule,
self.sim_freq, need_shift = self.need_shift,
database = self.config.get('dbconfig', dbaccess.dbconfig)['database'])
else:
cnx = dbaccess.connect(**self.config.get('dbconfig', dbaccess.dbconfig))
if self.sim_freq == 'd':
df = dbaccess.load_daily_data_to_df(cnx, dbtable, ticker,
self.config['start_date'], self.config['end_date'], index_col='date',
field = field_id)
else:
minid_start = 1500
minid_end = 2114
if ticker in misc.night_session_markets:
minid_start = 300
df = dbaccess.load_min_data_to_df(cnx, dbtable, ticker, self.config['start_date'],
self.config['end_date'], minid_start, minid_end)
df['contract'] = ticker
if self.sim_freq == 'm':
df = misc.cleanup_mindata(df, ticker)
df.columns = [(prod, col) for col in df.columns]
self.data_store[prod] = df
def get_data(spotID, start, end, spot_table = 'spot_daily', name = None, index_col = 'date', fx_pair = None, field = 'spotID', args = None):
cnx = dbaccess.connect(**dbaccess.dbconfig)
if args:
args['start_date'] = start
args['end_date'] = end
df = misc.nearby(spotID, **args)
df = df.reset_index()
else:
df = dbaccess.load_daily_data_to_df(cnx, spot_table, spotID, start, end, index_col = None, field = field)
if isinstance(df[index_col][0], basestring):
if len(df[index_col][0])> 12:
df[index_col] = df[index_col].apply(lambda x: datetime.datetime.strptime(x, "%Y-%m-%d %H:%M:%S").date())
else:
df[index_col] = df[index_col].apply(lambda x: datetime.datetime.strptime(x, "%Y-%m-%d").date())
df = df.set_index(index_col)
if name:
col_name = name
else:
col_name = spotID
if field == 'ccy':
df = df[df.tenor=='0W']
data_field = 'rate'
elif field == 'spotID':
data_field = 'close'
elif field == 'instID':
data_field = 'close'
df = df[[data_field]]
df.rename(columns = {data_field: col_name}, inplace = True)
if fx_pair:
fx = fx_pair.split('/')
direction = misc.get_mkt_fxpair(fx[0], fx[1])
if direction < 0:
mkt_pair = '/'.join([fx[1],fx[0]])
else:
mkt_pair = fx_pair
fx = dbaccess.load_daily_data_to_df(cnx, 'fx_daily', mkt_pair, start, end, index_col = None, field = 'ccy')
fx = fx[fx['tenor']=='0W']
if isinstance(fx[index_col][0], basestring):
fx[index_col] = fx[index_col].apply(lambda x: datetime.datetime.strptime(x, "%Y-%m-%d %H:%M:%S").date())
fx = fx.set_index(index_col)
if direction >= 0:
df[col_name] = df[col_name]/fx['rate']
else:
df[col_name] = df[col_name]*fx['rate']
return df
def simlauncher_min(config_file):
sim_config = {}
with open(config_file, 'r') as fp:
sim_config = json.load(fp)
bktest_split = sim_config['sim_func'].split('.')
bktest_module = __import__(bktest_split[0])
run_sim = getattr(bktest_module, bktest_split[1])
dir_name = config_file.split('.')[0]
test_folder = get_bktest_folder()
file_prefix = test_folder + dir_name + os.path.sep
if not os.path.exists(file_prefix):
os.makedirs(file_prefix)
sim_list = sim_config['products']
config = {}
start_date = datetime.datetime.strptime(sim_config['start_date'],
'%Y%m%d').date()
config['start_date'] = start_date
end_date = datetime.datetime.strptime(sim_config['end_date'],
'%Y%m%d').date()
config['end_date'] = end_date
scen_dim = [len(sim_config[s]) for s in sim_config['scen_keys']]
outcol_list = ['asset', 'scenario'] + sim_config['scen_keys'] \
+ ['sharp_ratio', 'tot_pnl', 'std_pnl', 'num_days', \
'max_drawdown', 'max_dd_period', 'profit_dd_ratio', \
'all_profit', 'tot_cost', 'win_ratio', 'num_win', 'num_loss', \
'profit_per_win', 'profit_per_loss']
scenarios = [list(s) for s in np.ndindex(tuple(scen_dim))]
config.update(sim_config['config'])
config['pos_class'] = eval(sim_config['pos_class'])
if 'proc_func' in sim_config:
config['proc_func'] = eval(sim_config['proc_func'])
file_prefix = file_prefix + sim_config['sim_name']
if config['close_daily']:
file_prefix = file_prefix + 'daily_'
config['file_prefix'] = file_prefix
summary_df = pd.DataFrame()
fname = config['file_prefix'] + 'summary.csv'
if os.path.isfile(fname):
summary_df = pd.DataFrame.from_csv(fname)
for asset in sim_list:
file_prefix = config['file_prefix'] + '_' + asset + '_'
fname = file_prefix + 'stats.json'
output = {}
if os.path.isfile(fname):
with open(fname, 'r') as fp:
output = json.load(fp)
if len(output.keys()) < len(scenarios):
if asset in sim_start_dict:
start_date = max(sim_start_dict[asset], config['start_date'])
else:
start_date = config['start_date']
if 'offset' in sim_config:
config[
'offset'] = sim_config['offset'] * trade_offset_dict[asset]
else:
config['offset'] = trade_offset_dict[asset]
config['marginrate'] = (sim_margin_dict[asset],
sim_margin_dict[asset])
config['nearby'] = 1
config['rollrule'] = '-50b'
config['exit_min'] = 2112
config['no_trade_set'] = range(300, 301) + range(
1500, 1501) + range(2059, 2100)
if asset in ['cu', 'al', 'zn']:
config['nearby'] = 3
config['rollrule'] = '-1b'
elif asset in ['IF', 'IH', 'IC']:
config['rollrule'] = '-2b'
config['no_trade_set'] = range(1515, 1520) + range(2110, 2115)
elif asset in ['au', 'ag']:
config['rollrule'] = '-25b'
elif asset in ['TF', 'T']:
config['rollrule'] = '-20b'
config['no_trade_set'] = range(1515, 1520) + range(2110, 2115)
config['no_trade_set'] = []
nearby = config['nearby']
rollrule = config['rollrule']
if nearby > 0:
mdf = misc.nearby(asset,
nearby,
start_date,
end_date,
rollrule,
'm',
need_shift=True,
database='hist_data')
mdf = cleanup_mindata(mdf, asset)
if 'need_daily' in sim_config:
ddf = misc.nearby(asset,
nearby,
start_date,
end_date,
rollrule,
'd',
need_shift=True,
database='hist_data')
config['ddf'] = ddf
for ix, s in enumerate(scenarios):
fname1 = file_prefix + str(ix) + '_trades.csv'
fname2 = file_prefix + str(ix) + '_dailydata.csv'
if os.path.isfile(fname1) and os.path.isfile(fname2):
continue
for key, seq in zip(sim_config['scen_keys'], s):
config[key] = sim_config[key][seq]
df = mdf.copy(deep=True)
(res, closed_trades, ts) = run_sim(df, config)
res.update(dict(zip(sim_config['scen_keys'], s)))
res['asset'] = asset
output[ix] = res
print 'saving results for asset = %s, scen = %s' % (asset,
str(ix))
all_trades = {}
for i, tradepos in enumerate(closed_trades):
all_trades[i] = strat.tradepos2dict(tradepos)
trades = pd.DataFrame.from_dict(all_trades).T
trades.to_csv(fname1)
ts.to_csv(fname2)
fname = file_prefix + 'stats.json'
try:
with open(fname, 'w') as ofile:
json.dump(output, ofile)
except:
continue
res = pd.DataFrame.from_dict(output, orient='index')
res.index.name = 'scenario'
res = res.sort(columns=['sharp_ratio'], ascending=False)
res = res.reset_index()
res.set_index(['asset', 'scenario'])
out_res = res[outcol_list]
if len(summary_df) == 0:
summary_df = out_res[:10].copy(deep=True)
else:
summary_df = summary_df.append(out_res[:10])
fname = config['file_prefix'] + 'summary.csv'
summary_df.to_csv(fname)
return
def simlauncher_min(config_file):
sim_config = {}
with open(config_file, 'r') as fp:
sim_config = json.load(fp)
bktest_split = sim_config['sim_func'].split('.')
bktest_module = __import__(bktest_split[0])
run_sim = getattr(bktest_module, bktest_split[1])
dir_name = config_file.split('.')[0]
test_folder = get_bktest_folder()
file_prefix = test_folder + dir_name + os.path.sep
if not os.path.exists(file_prefix):
os.makedirs(file_prefix)
sim_list = sim_config['products']
config = {}
start_date = datetime.datetime.strptime(sim_config['start_date'], '%Y%m%d').date()
config['start_date'] = start_date
end_date = datetime.datetime.strptime(sim_config['end_date'], '%Y%m%d').date()
config['end_date'] = end_date
scen_dim = [ len(sim_config[s]) for s in sim_config['scen_keys']]
outcol_list = ['asset', 'scenario'] + sim_config['scen_keys'] \
+ ['sharp_ratio', 'tot_pnl', 'std_pnl', 'num_days', \
'max_drawdown', 'max_dd_period', 'profit_dd_ratio', \
'all_profit', 'tot_cost', 'win_ratio', 'num_win', 'num_loss', \
'profit_per_win', 'profit_per_loss']
scenarios = [list(s) for s in np.ndindex(tuple(scen_dim))]
config.update(sim_config['config'])
config['pos_class'] = eval(sim_config['pos_class'])
if 'proc_func' in sim_config:
config['proc_func'] = eval(sim_config['proc_func'])
file_prefix = file_prefix + sim_config['sim_name']
if config['close_daily']:
file_prefix = file_prefix + 'daily_'
config['file_prefix'] = file_prefix
summary_df = pd.DataFrame()
fname = config['file_prefix'] + 'summary.csv'
if os.path.isfile(fname):
summary_df = pd.DataFrame.from_csv(fname)
for asset in sim_list:
file_prefix = config['file_prefix'] + '_' + asset + '_'
fname = file_prefix + 'stats.json'
output = {}
if os.path.isfile(fname):
with open(fname, 'r') as fp:
output = json.load(fp)
if len(output.keys()) < len(scenarios):
if asset in sim_start_dict:
start_date = max(sim_start_dict[asset], config['start_date'])
else:
start_date = config['start_date']
if 'offset' in sim_config:
config['offset'] = sim_config['offset'] * trade_offset_dict[asset]
else:
config['offset'] = trade_offset_dict[asset]
config['marginrate'] = ( sim_margin_dict[asset], sim_margin_dict[asset])
config['nearby'] = 1
config['rollrule'] = '-50b'
config['exit_min'] = 2112
config['no_trade_set'] = range(300, 301) + range(1500, 1501) + range(2059, 2100)
if asset in ['cu', 'al', 'zn']:
config['nearby'] = 3
config['rollrule'] = '-1b'
elif asset in ['IF', 'IH', 'IC']:
config['rollrule'] = '-2b'
config['no_trade_set'] = range(1515, 1520) + range(2110, 2115)
elif asset in ['au', 'ag']:
config['rollrule'] = '-25b'
elif asset in ['TF', 'T']:
config['rollrule'] = '-20b'
config['no_trade_set'] = range(1515, 1520) + range(2110, 2115)
config['no_trade_set'] = []
nearby = config['nearby']
rollrule = config['rollrule']
if nearby > 0:
mdf = misc.nearby(asset, nearby, start_date, end_date, rollrule, 'm', need_shift=True)
mdf = cleanup_mindata(mdf, asset)
if 'need_daily' in sim_config:
ddf = misc.nearby(asset, nearby, start_date, end_date, rollrule, 'd', need_shift=True)
config['ddf'] = ddf
for ix, s in enumerate(scenarios):
fname1 = file_prefix + str(ix) + '_trades.csv'
fname2 = file_prefix + str(ix) + '_dailydata.csv'
if os.path.isfile(fname1) and os.path.isfile(fname2):
continue
for key, seq in zip(sim_config['scen_keys'], s):
config[key] = sim_config[key][seq]
df = mdf.copy(deep = True)
(res, closed_trades, ts) = run_sim( df, config)
res.update(dict(zip(sim_config['scen_keys'], s)))
res['asset'] = asset
output[ix] = res
print 'saving results for asset = %s, scen = %s' % (asset, str(ix))
all_trades = {}
for i, tradepos in enumerate(closed_trades):
all_trades[i] = strat.tradepos2dict(tradepos)
trades = pd.DataFrame.from_dict(all_trades).T
trades.to_csv(fname1)
ts.to_csv(fname2)
fname = file_prefix + 'stats.json'
try:
with open(fname, 'w') as ofile:
json.dump(output, ofile)
except:
continue
res = pd.DataFrame.from_dict(output, orient = 'index')
res.index.name = 'scenario'
res = res.sort(columns = ['sharp_ratio'], ascending=False)
res = res.reset_index()
res.set_index(['asset', 'scenario'])
out_res = res[outcol_list]
if len(summary_df) == 0:
summary_df = out_res[:10].copy(deep = True)
else:
summary_df = summary_df.append(out_res[:10])
fname = config['file_prefix'] + 'summary.csv'
summary_df.to_csv(fname)
return
def aberration( asset, start_date, end_date, freqs, windows, config):
nearby = config['nearby']
rollrule = config['rollrule']
file_prefix = config['file_prefix'] + '_' + asset + '_'
df = misc.nearby(asset, nearby, start_date, end_date, rollrule, 'm', need_shift=True)