/
mod.py
509 lines (445 loc) · 23.5 KB
/
mod.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
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
# -*- coding: utf-8 -*-
# 版权所有 2019 深圳米筐科技有限公司(下称“米筐科技”)
#
# 除非遵守当前许可,否则不得使用本软件。
#
# * 非商业用途(非商业用途指个人出于非商业目的使用本软件,或者高校、研究所等非营利机构出于教育、科研等目的使用本软件):
# 遵守 Apache License 2.0(下称“Apache 2.0 许可”),
# 您可以在以下位置获得 Apache 2.0 许可的副本:http://www.apache.org/licenses/LICENSE-2.0。
# 除非法律有要求或以书面形式达成协议,否则本软件分发时需保持当前许可“原样”不变,且不得附加任何条件。
#
# * 商业用途(商业用途指个人出于任何商业目的使用本软件,或者法人或其他组织出于任何目的使用本软件):
# 未经米筐科技授权,任何个人不得出于任何商业目的使用本软件(包括但不限于向第三方提供、销售、出租、出借、转让本软件、
# 本软件的衍生产品、引用或借鉴了本软件功能或源代码的产品或服务),任何法人或其他组织不得出于任何目的使用本软件,
# 否则米筐科技有权追究相应的知识产权侵权责任。
# 在此前提下,对本软件的使用同样需要遵守 Apache 2.0 许可,Apache 2.0 许可与本许可冲突之处,以本许可为准。
# 详细的授权流程,请联系 public@ricequant.com 获取。
import os
import pandas
import pickle
import jsonpickle
import datetime
from operator import attrgetter
from collections import defaultdict
from typing import Dict, Optional, List, Tuple, Union, Iterable
import numpy as np
import pandas as pd
from rqrisk import Risk, WEEKLY, MONTHLY
from rqalpha.const import EXIT_CODE, DEFAULT_ACCOUNT_TYPE, INSTRUMENT_TYPE, POSITION_DIRECTION
from rqalpha.core.events import EVENT
from rqalpha.interface import AbstractMod, AbstractPosition
from rqalpha.utils.i18n import gettext as _
from rqalpha.utils import INST_TYPE_IN_STOCK_ACCOUNT
from rqalpha.utils.logger import user_system_log
from rqalpha.const import DAYS_CNT
from rqalpha.api import export_as_api
from .plot.consts import DefaultPlot, PLOT_TEMPLATE
from .plot.utils import max_ddd as _max_ddd
from .plot_store import PlotStore
def _get_yearly_risk_free_rates(
data_proxy, start_date: datetime.date, end_date: datetime.date
) -> Iterable[Tuple[int, float]]:
while start_date <= end_date:
year = start_date.year
yield year, data_proxy.get_risk_free_rate(start_date, min(end_date, datetime.date(year, 12, 31)))
start_date = datetime.date(year + 1, 1, 1)
class AnalyserMod(AbstractMod):
def __init__(self):
self._env = None
self._mod_config = None
self._enabled = False
self._orders = []
self._trades = []
self._total_portfolios = []
self._total_benchmark_portfolios = []
self._sub_accounts = defaultdict(list)
self._positions = defaultdict(list)
self._daily_pnl = []
self._benchmark_daily_returns = []
self._portfolio_daily_returns = []
self._benchmark: Optional[List[Tuple[str, float]]] = None
self._plot_store = None
def get_state(self):
return jsonpickle.dumps({
'benchmark_daily_returns': [float(v) for v in self._benchmark_daily_returns],
'portfolio_daily_returns': [float(v) for v in self._portfolio_daily_returns],
'total_portfolios': self._total_portfolios,
'total_benchmark_portfolios': self._total_benchmark_portfolios,
'sub_accounts': self._sub_accounts,
'positions': self._positions,
'orders': self._orders,
'trades': self._trades,
'daily_pnl': self._daily_pnl,
}).encode('utf-8')
def set_state(self, state):
value = jsonpickle.loads(state.decode('utf-8'))
self._benchmark_daily_returns = value['benchmark_daily_returns']
self._portfolio_daily_returns = value["portfolio_daily_returns"]
self._total_portfolios = value['total_portfolios']
self._total_benchmark_portfolios = value["total_benchmark_portfolios"]
self._sub_accounts = value['sub_accounts']
self._positions = value["positions"]
self._orders = value['orders']
self._trades = value["trades"]
self._daily_pnl = value.get("daily_pnl", [])
def start_up(self, env, mod_config):
self._env = env
self._mod_config = mod_config
self._enabled = (
mod_config.record or mod_config.plot or mod_config.output_file or
mod_config.plot_save_file or mod_config.report_save_path or mod_config.benchmark
)
if self._enabled:
env.event_bus.add_listener(EVENT.POST_SYSTEM_INIT, self._subscribe_events)
if not mod_config.benchmark:
if getattr(env.config.base, "benchmark", None):
user_system_log.warning(
_("config 'base.benchmark' is deprecated, use 'mod.sys_analyser.benchmark' instead")
)
mod_config.benchmark = getattr(env.config.base, "benchmark")
if mod_config.benchmark:
self._benchmark = self._parse_benchmark(mod_config.benchmark)
self._plot_store = PlotStore(env)
export_as_api(self._plot_store.plot)
def get_benchmark_daily_returns(self):
if self._benchmark is None:
return np.nan
daily_return_list = []
weights = 0
for benchmark in self._benchmark:
bar = self._env.data_proxy.get_bar(benchmark[0], self._env.calendar_dt, '1d')
if bar.close != bar.close:
daily_return_list.append((0.0, benchmark[1]))
else:
daily_return_list.append((bar.close / bar.prev_close - 1.0, benchmark[1]))
weights += benchmark[1]
return sum([daily[0] * daily[1] / weights for daily in daily_return_list])
def _subscribe_events(self, _):
self._env.event_bus.add_listener(EVENT.TRADE, self._collect_trade)
self._env.event_bus.add_listener(EVENT.ORDER_CREATION_PASS, self._collect_order)
self._env.event_bus.add_listener(EVENT.POST_SETTLEMENT, self._collect_daily)
def _collect_trade(self, event):
self._trades.append(self._to_trade_record(event.trade))
def _collect_order(self, event):
self._orders.append(event.order)
def _collect_daily(self, _):
date = self._env.calendar_dt.date()
portfolio = self._env.portfolio
self._portfolio_daily_returns.append(portfolio.daily_returns)
self._total_portfolios.append(self._to_portfolio_record(date, portfolio))
self._benchmark_daily_returns.append(self.get_benchmark_daily_returns())
self._total_benchmark_portfolios.append({
"date": date,
"unit_net_value": (np.array(self._benchmark_daily_returns) + 1).prod()
})
self._daily_pnl.append(portfolio.daily_pnl)
for account_type, account in self._env.portfolio.accounts.items():
self._sub_accounts[account_type].append(self._to_account_record(date, account))
pos_dict = {}
for pos in account.get_positions():
pos_dict.setdefault(pos.order_book_id, {})[pos.direction] = pos
for order_book_id, pos in pos_dict.items():
self._positions[account_type].append(self._to_position_record(
date, order_book_id, pos.get(POSITION_DIRECTION.LONG), pos.get(POSITION_DIRECTION.SHORT)
))
def _symbol(self, order_book_id):
return self._env.data_proxy.instrument(order_book_id).symbol
@staticmethod
def _parse_benchmark(benchmarks):
# type: (Union[str, Dict]) -> List[Tuple[str, float]]
# --benchmark 000001.XSHE:1000,IF1701:-1
result = []
if not isinstance(benchmarks, str):
# 字典
for order_book_id, weight in benchmarks.items():
result.append((order_book_id, float(weight)))
return result
benchmark_list = benchmarks.split(',')
if len(benchmark_list) == 1:
if len(benchmark_list[0].split(':')) > 1:
result.append((benchmark_list[0].split(':')[0], 1.0))
return result
result.append((benchmark_list[0], 1.0))
return result
for s in benchmark_list:
try:
order_book_id, weight = s.split(':')
except ValueError:
raise RuntimeError(
_(u"invalid init benchmark {}, should be in format 'order_book_id:weight'").format(s))
try:
result.append((order_book_id, float(weight)))
except ValueError:
raise RuntimeError(_(u"invalid weight for instrument {order_book_id}: {weight}").format(
order_book_id=order_book_id, weight=weight))
return result
@staticmethod
def _safe_convert(value, ndigits=4):
try:
return round(float(value), ndigits)
except TypeError:
return value
def _to_portfolio_record(self, date, portfolio):
return {
'date': date,
'cash': self._safe_convert(portfolio.cash),
'total_value': self._safe_convert(portfolio.total_value),
'market_value': self._safe_convert(portfolio.market_value),
'unit_net_value': self._safe_convert(portfolio.unit_net_value, 6),
'units': portfolio.units,
'static_unit_net_value': self._safe_convert(portfolio.static_unit_net_value),
}
ACCOUNT_FIELDS_MAP = {
DEFAULT_ACCOUNT_TYPE.STOCK: [],
DEFAULT_ACCOUNT_TYPE.FUTURE: ['position_pnl', 'trading_pnl', 'daily_pnl', 'margin'],
DEFAULT_ACCOUNT_TYPE.BOND: [],
}
def _to_account_record(self, date, account):
data = {
'date': date,
'cash': self._safe_convert(account.cash),
'transaction_cost': self._safe_convert(account.transaction_cost),
'market_value': self._safe_convert(account.market_value),
'total_value': self._safe_convert(account.total_value),
}
for f in self.ACCOUNT_FIELDS_MAP[account.type]:
data[f] = self._safe_convert(getattr(account, f))
return data
LONG_ONLY_INS_TYPE = INST_TYPE_IN_STOCK_ACCOUNT + [INSTRUMENT_TYPE.CONVERTIBLE, INSTRUMENT_TYPE.BOND]
def _to_position_record(self, date, order_book_id, long, short):
# type: (datetime.date, str, AbstractPosition, AbstractPosition) -> Dict
instrument = self._env.data_proxy.instrument(order_book_id)
data = {
'order_book_id': order_book_id,
'symbol': self._symbol(order_book_id),
'date': date,
}
if instrument.type in self.LONG_ONLY_INS_TYPE + [INSTRUMENT_TYPE.REPO]:
for field in ['quantity', 'last_price', 'avg_price', 'market_value']:
data[field] = self._safe_convert(getattr(long, field, None))
else:
position = long or short
if position:
for field in ['margin', 'contract_multiplier', 'last_price']:
data[field] = self._safe_convert(getattr(position, field))
direction_pos_iter = ((pos.direction, pos) for pos in (long, short) if pos)
for direction_prefix, pos in direction_pos_iter:
data[direction_prefix + "_pnl"] = self._safe_convert(getattr(pos, "pnl", None))
data[direction_prefix + "_margin"] = self._safe_convert(pos.margin)
data[direction_prefix + "_quantity"] = self._safe_convert(pos.quantity)
data[direction_prefix + "_avg_open_price"] = self._safe_convert(getattr(pos, "avg_price", None))
return data
def _to_trade_record(self, trade):
return {
'datetime': trade.datetime.strftime("%Y-%m-%d %H:%M:%S"),
'trading_datetime': trade.trading_datetime.strftime("%Y-%m-%d %H:%M:%S"),
'order_book_id': trade.order_book_id,
'symbol': self._symbol(trade.order_book_id),
'side': trade.side.name,
'position_effect': trade.position_effect.name,
'exec_id': trade.exec_id,
'tax': trade.tax,
'commission': trade.commission,
'last_quantity': trade.last_quantity,
'last_price': self._safe_convert(trade.last_price),
'order_id': trade.order_id,
'transaction_cost': trade.transaction_cost,
}
def tear_down(self, code, exception=None):
if code != EXIT_CODE.EXIT_SUCCESS or not self._enabled:
return
# 当 PRE_SETTLEMENT 事件没有被触发当时候,self._total_portfolio 为空list
if len(self._total_portfolios) == 0:
return
strategy_name = os.path.basename(self._env.config.base.strategy_file).split(".")[0]
data_proxy = self._env.data_proxy
start_date, end_date = attrgetter("start_date", "end_date")(self._env.config.base)
summary = {
'strategy_name': strategy_name,
'start_date': start_date.strftime('%Y-%m-%d'),
'end_date': end_date.strftime('%Y-%m-%d'),
'strategy_file': self._env.config.base.strategy_file,
'run_type': self._env.config.base.run_type.value,
"starting_cash": ",".join(f"{t}:{c}" for t, c in self._env.config.base.accounts.items())
}
for account_type, starting_cash in self._env.config.base.accounts.items():
summary[account_type] = starting_cash
if self._benchmark:
if len(self._benchmark) == 1:
benchmark_obid, _ = self._benchmark[0]
summary["benchmark"] = benchmark_obid
summary["benchmark_symbol"] = self._env.data_proxy.instrument(benchmark_obid).symbol
else:
summary["benchmark"] = ",".join(f"{o}:{w}" for o, w in self._benchmark)
summary["benchmark_symbol"] = ",".join(f"{self._env.data_proxy.instrument(o).symbol}:{w}" for o, w in self._benchmark)
risk_free_rate = data_proxy.get_risk_free_rate(self._env.config.base.start_date, self._env.config.base.end_date)
risk = Risk(
np.array(self._portfolio_daily_returns), np.array(self._benchmark_daily_returns), risk_free_rate
)
summary.update({
'alpha': risk.alpha,
'beta': risk.beta,
'sharpe': risk.sharpe,
'excess_sharpe': risk.excess_sharpe,
'information_ratio': risk.information_ratio,
'downside_risk': risk.annual_downside_risk,
'tracking_error': risk.annual_tracking_error,
'sortino': risk.sortino,
'volatility': risk.annual_volatility,
'excess_volatility': risk.excess_annual_volatility,
'excess_annual_volatility': risk.excess_annual_volatility,
'max_drawdown': risk.max_drawdown,
'excess_max_drawdown': risk.excess_max_drawdown,
'excess_returns': risk.excess_return_rate,
'excess_annual_returns': risk.excess_annual_return,
'var': risk.var,
"win_rate": risk.win_rate,
"excess_win_rate": risk.excess_win_rate,
})
# 盈亏比
daily_pnl = np.array(self._daily_pnl)
profit, loss = daily_pnl[daily_pnl > 0], daily_pnl[daily_pnl < 0]
profit, loss = profit.mean() if len(profit) else 0, loss.mean() if len(loss) else 0
summary.update({
"profit_loss_rate": np.abs(profit / loss) if loss else np.nan
})
summary.update({
'total_value': self._env.portfolio.total_value,
'cash': self._env.portfolio.cash,
'total_returns': self._env.portfolio.total_returns,
'annualized_returns': self._env.portfolio.annualized_returns,
'unit_net_value': self._env.portfolio.unit_net_value,
'units': self._env.portfolio.units,
})
if self._benchmark:
benchmark_total_returns = (np.array(self._benchmark_daily_returns) + 1.0).prod() - 1.0
summary['benchmark_total_returns'] = benchmark_total_returns
date_count = len(self._benchmark_daily_returns)
benchmark_annualized_returns = (benchmark_total_returns + 1) ** (DAYS_CNT.TRADING_DAYS_A_YEAR / date_count) - 1
summary['benchmark_annualized_returns'] = benchmark_annualized_returns
# 新增一个超额累计收益
summary['excess_cum_returns'] = summary["total_returns"] - summary["benchmark_total_returns"]
trades = pd.DataFrame(self._trades)
if 'datetime' in trades.columns:
trades = trades.set_index(pd.DatetimeIndex(trades['datetime']))
df = pd.DataFrame(self._total_portfolios)
df['date'] = pd.to_datetime(df['date'])
total_portfolios = df.set_index('date').sort_index()
df.index = df['date']
weekly_nav = df.resample("W").last().set_index("date").unit_net_value.dropna()
monthly_nav = df.resample("M").last().set_index("date").unit_net_value.dropna()
weekly_returns = (weekly_nav / weekly_nav.shift(1).fillna(1)).fillna(0) - 1
monthly_returns = (monthly_nav / monthly_nav.shift(1).fillna(1)).fillna(0) - 1
# 最长回撤持续期
max_ddd = _max_ddd(total_portfolios.unit_net_value.values, total_portfolios.index)
summary["max_drawdown_duration"] = max_ddd
summary.update({
"max_drawdown_duration_start_date": str(max_ddd.start_date),
"max_drawdown_duration_end_date": str(max_ddd.end_date),
"max_drawdown_duration_days": (max_ddd.end_date - max_ddd.start_date).days,
})
result_dict = {
'summary': summary,
'trades': trades,
'portfolio': total_portfolios,
}
if not trades.empty and all(
trades.order_book_id.str.endswith(".XSHE") | trades.order_book_id.str.endswith(".XSHG")
):
# 策略仅交易股票、指数、场内基金等品种时才计算换手率
trades_values = trades.last_price * trades.last_quantity
market_values = total_portfolios.market_value
summary["turnover"] = trades_values.sum() / market_values.mean() / 2
avg_daily_turnover = (trades_values.groupby(trades.index.date).sum() / market_values / 2)
with pd.option_context('mode.use_inf_as_na', True):
summary["avg_daily_turnover"] = avg_daily_turnover.fillna(0).mean()
else:
summary["turnover"] = np.nan
if self._benchmark:
df = pd.DataFrame(self._total_benchmark_portfolios)
df['date'] = pd.to_datetime(df['date'])
benchmark_portfolios = df.set_index('date').sort_index()
df.index = df['date']
weekly_b_nav = df.resample("W").last().set_index("date").unit_net_value.dropna()
monthly_b_nav = df.resample("M").last().set_index("date").unit_net_value.dropna()
weekly_b_returns = (weekly_b_nav / weekly_b_nav.shift(1).fillna(1)).fillna(0) - 1
monthly_b_returns = (monthly_b_nav / monthly_b_nav.shift(1).fillna(1)).fillna(0) - 1
result_dict['benchmark_portfolio'] = benchmark_portfolios
# 超额收益最长回撤持续期
ex_returns = total_portfolios.unit_net_value - benchmark_portfolios.unit_net_value
max_ddd = _max_ddd(ex_returns + 1, total_portfolios.index)
result_dict["summary"]["excess_max_drawdown_duration"] = max_ddd
result_dict["summary"]["excess_max_drawdown_duration_start_date"] = str(max_ddd.start_date)
result_dict["summary"]["excess_max_drawdown_duration_end_date"] = str(max_ddd.end_date)
result_dict["summary"]["excess_max_drawdown_duration_days"] = (max_ddd.end_date - max_ddd.start_date).days
else:
weekly_b_returns = pandas.Series(index=weekly_returns.index)
monthly_b_returns = pandas.Series(index=monthly_returns.index)
# 周度风险指标
weekly_risk = Risk(weekly_returns, weekly_b_returns, risk_free_rate, WEEKLY)
summary.update({
"weekly_alpha": weekly_risk.alpha,
"weekly_beta": weekly_risk.beta,
"weekly_sharpe": weekly_risk.sharpe,
"weekly_sortino": weekly_risk.sortino,
"weekly_information_ratio": weekly_risk.information_ratio,
"weekly_tracking_error": weekly_risk.annual_tracking_error,
"weekly_max_drawdown": weekly_risk.max_drawdown,
"weekly_win_rate": weekly_risk.win_rate,
"weekly_volatility": weekly_risk.annual_volatility,
"weekly_ulcer_index": weekly_risk.ulcer_index,
"weekly_ulcer_performance_index": weekly_risk.ulcer_performance_index,
})
# 月度风险指标
monthly_risk = Risk(monthly_returns, monthly_b_returns, risk_free_rate, MONTHLY)
summary.update({
"monthly_sharpe": monthly_risk.sharpe,
"monthly_volatility": monthly_risk.annual_volatility,
"monthly_excess_win_rate": monthly_risk.excess_win_rate,
})
if self._benchmark:
summary.update({
"weekly_excess_ulcer_index": weekly_risk.excess_ulcer_index,
"weekly_excess_ulcer_performance_index": weekly_risk.excess_ulcer_performance_index,
})
plots = self._plot_store.get_plots()
if plots:
plots_items = defaultdict(dict)
for series_name, value_dict in plots.items():
for date, value in value_dict.items():
plots_items[date][series_name] = value
plots_items[date]["date"] = date
df = pd.DataFrame([dict_data for date, dict_data in plots_items.items()])
df["date"] = pd.to_datetime(df["date"])
df = df.set_index("date").sort_index()
result_dict["plots"] = df
for account_type, account in self._env.portfolio.accounts.items():
account_name = account_type.lower()
portfolios_list = self._sub_accounts[account_type]
df = pd.DataFrame(portfolios_list)
df["date"] = pd.to_datetime(df["date"])
df = df.set_index("date").sort_index()
result_dict["{}_account".format(account_name)] = df
positions_list = self._positions[account_type]
df = pd.DataFrame(positions_list)
if "date" in df.columns:
df["date"] = pd.to_datetime(df["date"])
df = df.set_index("date").sort_index()
result_dict["{}_positions".format(account_name)] = df
result_dict["yearly_risk_free_rates"] = dict(_get_yearly_risk_free_rates(data_proxy, start_date, end_date))
if self._mod_config.output_file:
with open(self._mod_config.output_file, 'wb') as f:
pickle.dump(result_dict, f)
if self._mod_config.report_save_path:
from .report import generate_report
generate_report(result_dict, self._mod_config.report_save_path)
_plot = True if self._mod_config.plot else False
if _plot or self._mod_config.plot_save_file:
from .plot import plot_result
plot_config = self._mod_config.plot_config
_plot_template_cls = PLOT_TEMPLATE.get(self._mod_config.plot, DefaultPlot)
plot_result(
result_dict, self._mod_config.plot, self._mod_config.plot_save_file,
plot_config.weekly_indicators, plot_config.open_close_points, _plot_template_cls
)
return result_dict