forked from harvey1673/pyktrader
-
Notifications
You must be signed in to change notification settings - Fork 3
/
backtest.py
312 lines (299 loc) · 13.7 KB
/
backtest.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
import datetime
import os
import sys
import json
import pandas as pd
import numpy as np
import data_handler as dh
import strategy as strat
import misc
import platform
sim_margin_dict = { 'au': 0.06, 'ag': 0.08, 'cu': 0.07, 'al':0.05,
'zn': 0.06, 'rb': 0.06, 'ru': 0.12, 'a': 0.05,
'm': 0.05, 'RM': 0.05, 'y' : 0.05, 'p': 0.05,
'c': 0.05, 'CF': 0.05, 'i' : 0.05, 'j': 0.05,
'jm': 0.05, 'pp': 0.05, 'l' : 0.05, 'SR': 0.06,
'TA': 0.06, 'TC': 0.05, 'ME': 0.06, 'IF': 0.1,
'jd': 0.06, 'ni': 0.07, 'IC': 0.1, 'ZC': 0.05,
'IH': 0.01, 'FG': 0.05, 'TF':0.015, 'OI': 0.05,
'T': 0.015, 'MA': 0.06, 'cs': 0.05, 'bu': 0.07,
'sn': 0.05, 'v': 0.05 }
sim_start_dict = { 'c': datetime.date(2008,10,1), 'm': datetime.date(2010,10,1),
'y': datetime.date(2010,10,1), 'l': datetime.date(2010,10,1), 'rb':datetime.date(2010,10,1),
'p': datetime.date(2010,10,1), 'cu':datetime.date(2010,10,1), 'al':datetime.date(2010,10,1),
'zn':datetime.date(2010,10,1), 'au':datetime.date(2010,10,1), 'v': datetime.date(2010,10,1),
'a': datetime.date(2010,10,1), 'ru':datetime.date(2010,10,1), 'ag':datetime.date(2012,7,1),
'i': datetime.date(2013,11,26), 'j': datetime.date(2011,6,1), 'jm':datetime.date(2013,5,1),
'ME':datetime.date(2012,2,1), 'CF':datetime.date(2012,6,1), 'TA':datetime.date(2012,6,1),
'PM':datetime.date(2012,10,1), 'RM':datetime.date(2013,2,1), 'SR':datetime.date(2013,2,1),
'FG':datetime.date(2013,2,1), 'OI':datetime.date(2013,6,1), 'RI':datetime.date(2013,6,1),
'TC':datetime.date(2013,10,1), 'WH':datetime.date(2014,2,1), 'pp':datetime.date(2014,4,1),
'IF':datetime.date(2010,7,1), 'MA':datetime.date(2015,7,1), 'TF':datetime.date(2014,4,1),
'IH':datetime.date(2015,5,1), 'IC':datetime.date(2015,5,1), 'cs':datetime.date(2015,2,2),
'jd':datetime.date(2014,6,1), 'ni':datetime.date(2015,6,1), 'sn':datetime.date(2015,6,1),
'ZC':datetime.date(2015,12,1),
}
trade_offset_dict = {
'au': 0.05, 'ag': 1, 'cu': 10, 'al':5,
'zn': 5, 'rb': 1, 'ru': 5, 'a': 1,
'm': 1, 'RM': 1, 'y' : 2, 'p': 2,
'c': 1, 'CF': 5, 'i' : 0.5, 'j': 0.5,
'jm': 0.5, 'pp': 1, 'l' : 5, 'SR': 1,
'TA': 2, 'TC': 0.2, 'ME': 1, 'IF': 0.4,
'jd': 1, 'ni': 10, 'IC': 1.0,
'IH': 0.4, 'FG': 1, 'TF':0.005, 'OI': 2,
'T': 0.005, 'MA': 1, 'cs': 1, 'bu': 1,
'sn': 10, 'v': 5,
}
def get_bktest_folder():
folder = ''
system = platform.system()
if system == 'Linux':
folder = '/home/harvey/dev/pyctp2/results/'
elif system == 'Windows':
folder = 'C:\\dev\\pyktlib\\pyktrader2\\results\\'
return folder
def get_asset_tradehrs(asset):
exch = 'SHFE'
for ex in misc.product_code:
if asset in misc.product_code[ex]:
exch = ex
break
hrs = [(1500, 1615), (1630, 1730), (1930, 2100)]
if exch in ['SSE', 'SZE']:
hrs = [(1530, 1730), (1900, 2100)]
elif exch == 'CFFEX':
hrs = [(1515, 1730), (1900, 2115)]
else:
if asset in misc.night_session_markets:
night_idx = misc.night_session_markets[asset]
hrs = [misc.night_trading_hrs[night_idx]] + hrs
return hrs
def cleanup_mindata(df, asset):
cond = None
tradehrs = get_asset_tradehrs(asset)
for idx, hrs in enumerate(tradehrs):
if idx == 0:
cond = (df.min_id>= tradehrs[idx][0]) & (df.min_id < tradehrs[idx][1])
else:
cond = cond | (df.min_id>= tradehrs[idx][0]) & (df.min_id < tradehrs[idx][1])
if asset in ['a', 'b', 'p', 'y', 'm', 'i', 'j', 'm']:
cond = cond | ((df.index < datetime.datetime(2015, 5, 12, 15, 0, 0)) & (df.min_id>=300) & (df.min_id<830))
df = df.ix[cond]
df = df[(df.close > 0) & (df.high > 0) & (df.open > 0) & (df.low > 0)]
return df
def get_pnl_stats(df, start_capital, marginrate, freq):
df['pnl'] = df['pos'].shift(1)*(df['close'] - df['close'].shift(1)).fillna(0.0)
df['margin'] = pd.concat([df.pos*marginrate[0]*df.close, -df.pos*marginrate[1]*df.close], join='outer', axis=1).max(1)
if freq == 'm':
daily_pnl = pd.Series(df['pnl']).resample('1d',how='sum').dropna()
daily_margin = pd.Series(df['margin']).resample('1d',how='last').dropna()
daily_cost = pd.Series(df['cost']).resample('1d',how='sum').dropna()
else:
daily_pnl = pd.Series(df['pnl'])
daily_margin = pd.Series(df['margin'])
daily_cost = pd.Series(df['cost'])
daily_pnl.name = 'daily_pnl'
daily_margin.name = 'daily_margin'
daily_cost.name = 'daily_cost'
cum_pnl = pd.Series(daily_pnl.cumsum() + daily_cost.cumsum() + start_capital, name = 'cum_pnl')
available = cum_pnl - daily_margin
res = {}
res['avg_pnl'] = daily_pnl.mean()
res['std_pnl'] = daily_pnl.std()
res['tot_pnl'] = daily_pnl.sum()
res['tot_cost'] = daily_cost.sum()
res['num_days'] = len(daily_pnl)
res['sharp_ratio'] = res['avg_pnl']/res['std_pnl']*np.sqrt(252.0)
max_dd, max_dur = max_drawdown(cum_pnl)
res['max_margin'] = daily_margin.max()
res['min_avail'] = available.min()
res['max_drawdown'] = max_dd
res['max_dd_period'] = max_dur
if abs(max_dd) > 0:
res['profit_dd_ratio'] = res['tot_pnl']/abs(max_dd)
else:
res['profit_dd_ratio'] = 0
ts = pd.concat([cum_pnl, daily_margin, daily_cost], join='outer', axis=1)
return res, ts
def get_trade_stats(trade_list):
res = {}
res['n_trades'] = len(trade_list)
res['all_profit'] = sum([trade.profit for trade in trade_list])
res['win_profit'] = sum([trade.profit for trade in trade_list if trade.profit>0])
res['loss_profit'] = sum([trade.profit for trade in trade_list if trade.profit<0])
sorted_profit = sorted([trade.profit for trade in trade_list])
res['largest_profit'] = sorted_profit[-1]
res['second largest'] = sorted_profit[-2]
res['third_profit'] = sorted_profit[-3]
res['largest_loss'] = sorted_profit[0]
res['second_loss'] = sorted_profit[1]
res['third_loss'] = sorted_profit[2]
res['num_win'] = len([trade.profit for trade in trade_list if trade.profit>0])
res['num_loss'] = len([trade.profit for trade in trade_list if trade.profit<0])
res['win_ratio'] = 0
if res['n_trades'] > 0:
res['win_ratio'] = float(res['num_win'])/float(res['n_trades'])
res['profit_per_win'] = 0
if res['num_win'] > 0:
res['profit_per_win'] = res['win_profit']/float(res['num_win'])
res['profit_per_loss'] = 0
if res['num_loss'] > 0:
res['profit_per_loss'] = res['loss_profit']/float(res['num_loss'])
return res
def create_drawdowns(ts):
"""
Calculate the largest peak-to-trough drawdown of the PnL curve
as well as the duration of the drawdown. Requires that the
pnl_returns is a pandas Series.
Parameters:
pnl - A pandas Series representing period percentage returns.
Returns:
drawdown, duration - Highest peak-to-trough drawdown and duration.
"""
# Calculate the cumulative returns curve
# and set up the High Water Mark
# Then create the drawdown and duration series
ts_idx = ts.index
drawdown = pd.Series(index = ts_idx)
duration = pd.Series(index = ts_idx)
hwm = pd.Series([0]*len(ts), index = ts_idx)
last_t = ts_idx[0]
# Loop over the index range
for idx, t in enumerate(ts_idx):
if idx > 0:
cur_hwm = max(hwm[last_t], ts_idx[idx])
hwm[t] = cur_hwm
drawdown[t]= hwm[t] - ts[t]
duration[t]= 0 if drawdown[t] == 0 else duration[last_t] + 1
last_t = t
return drawdown.max(), duration.max()
def max_drawdown(ts):
i = np.argmax(np.maximum.accumulate(ts)-ts)
j = np.argmax(ts[:i])
max_dd = ts[i] - ts[j]
max_duration = (i - j).days
return max_dd, max_duration
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
if __name__=="__main__":
args = sys.argv[1:]
if len(args) < 1:
print "need to input a file name for simulation"
else:
simlauncher_min(args[0])
pass