def strategyRunner(userStrategy,
strategyParameters=(),
initialCapital=100000,
symbolList=['600000.XSHG'],
startDate=dt.datetime(2015, 9, 1),
endDate=dt.datetime(2015, 9, 15),
benchmark=None,
refreshRate=1,
saveFile=False,
plot=True,
logLevel='critical',
portfolioType=PortfolioType.FullNotional,
**kwargs):
u"""
回测用运行器
:param userStrategy: 用户自定义策略
:param strategyParameters: 用户自定义策略所用参数，默认为None
:param initialCapital: 初始资金，默认为100000
:param symbolList: 证券代码，默认为600000.xshg
:param startDate: 回测起始日，默认为2015年9月1日
:param endDate: 回测结束日，默认为2015年9月15日
:param dataSource: 行情数据源，默认为DataSource.DXDataCenter
:param benchmark: 比较基准，比如指数。默认为None
:param refreshRate: bar的调用频率，默认为0
:param saveFile: 是否保存表现的数据为本地文件，默认为False
:param plot: 是否绘制表现图，默认为True
:param logLevel: 日志的等级
:param portfolioType: 收益计算的方法
:param kwargs: 其他需要的关键字参数
:return: dict
"""
logger = CustomLogger(logLevel)
try:
dataSource = kwargs['dataSource']
Settings.set_source(dataSource)
except KeyError:
dataSource = Settings.data_source
if dataSource == DataSource.CSV:
dataHandler = HistoricalCSVDataHandler(csvDir=kwargs['csvDir'],
symbolList=symbolList,
logger=logger)
elif dataSource == DataSource.DataYes:
try:
token = kwargs['token']
except KeyError:
token = None
dataHandler = DataYesMarketDataHandler(token=token,
symbolList=symbolList,
startDate=startDate,
endDate=endDate,
benchmark=benchmark,
logger=logger)
elif dataSource == DataSource.DXDataCenter:
try:
freq = kwargs['freq']
except KeyError:
freq = 0
logger.info(
"No `freq` keyword arguments found. using default value as freq=0"
)
dataHandler = DXDataCenter(symbolList=symbolList,
startDate=startDate,
endDate=endDate,
freq=freq,
benchmark=benchmark,
logger=logger)
elif dataSource == DataSource.YAHOO:
dataHandler = YaHooDataProvider(symbolList=symbolList,
startDate=startDate,
endDate=endDate,
logger=logger)
elif dataSource == DataSource.WIND:
try:
freq = kwargs['freq']
except KeyError:
freq = 'D'
logger.info(
"No `freq` keyword arguments found. using default value as freq=D"
)
try:
priceAdj = kwargs['priceAdj']
except KeyError:
priceAdj = '0'
logger.info(
"No `priceAdj` keyword arguments found. using default value as priceAdj=0"
)
dataHandler = WindMarketDataHandler(symbolList=symbolList,
startDate=startDate,
endDate=endDate,
freq=freq,
priceAdj=priceAdj,
benchmark=benchmark,
logger=logger)
backtest = Backtest(initialCapital,
0.0,
dataHandler,
SimulatedExecutionHandler,
Portfolio,
userStrategy,
logger,
benchmark,
refreshRate,
plot=plot,
portfolioType=portfolioType,
strategyParameters=strategyParameters)
equityCurve, \
orderBook, \
filledBook, \
perf_metric, \
perf_df, \
rollingRisk, \
aggregated_positions, \
transactions, \
turnover_rate, \
info_view \
= backtest.simulateTrading()
# save to a excel file
if saveFile:
if os.path.isdir('performance/'):
shutil.rmtree('performance/')
os.mkdir('performance/')
logger.info("Strategy performance is now saving to local files...")
perf_metric.to_csv('performance/perf_metrics.csv', float_format='%.4f')
perf_df.to_csv('performance/perf_series.csv', float_format='%.4f')
if benchmark is not None and rollingRisk is not None:
rollingRisk.to_csv('performance/rollingRisk.csv',
float_format='%.4f')
equityCurve.to_csv('performance/equity_curve.csv', float_format='%.4f')
orderBook.to_csv('performance/order_book.csv', float_format='%.4f')
filledBook.to_csv('performance/filled_book.csv', float_format='%.4f')
aggregated_positions.to_csv('performance/aggregated_positions.csv',
float_format='%.4f')
turnover_rate.to_csv('performance/turnover_rate.csv',
float_format='%.4f')
transactions.to_csv('performance/transactions.csv',
float_format='%.4f')
info_view.to_csv('performance/strategy_info.csv', float_format='%.4f')
logger.info("Performance saving is finished!")
return {
'equity_curve': equityCurve,
'order_book': orderBook,
'filled_book': filledBook,
'perf_metric': perf_metric,
'perf_series': perf_df,
'aggregated_positions': aggregated_positions,
'transactions': transactions,
'turnover_rate': turnover_rate,
'user_info': info_view
}
def strategyRunner(userStrategy,
strategyParameters=(),
initialCapital=100000,
symbolList=['600000.XSHG'],
startDate=dt.datetime(2015, 9, 1),
endDate=dt.datetime(2015, 9, 15),
dataSource=DataSource.DXDataCenter,
benchmark=None,
refreshRate=1,
saveFile=False,
plot=True,
logLevel='critical',
portfolioType=PortfolioType.FullNotional,
**kwargs):
logger = CustomLogger(logLevel)
if dataSource == DataSource.CSV:
dataHandler = HistoricalCSVDataHandler(csvDir=kwargs['csvDir'],
symbolList=symbolList,
logger=logger)
elif dataSource == DataSource.DataYes:
try:
token = kwargs['token']
except KeyError:
token = None
dataHandler = DataYesMarketDataHandler(token=token,
symbolList=symbolList,
startDate=startDate,
endDate=endDate,
benchmark=benchmark,
logger=logger)
elif dataSource == DataSource.DXDataCenter:
try:
freq = kwargs['freq']
except KeyError:
freq = 0
dataHandler = DXDataCenter(symbolList=symbolList,
startDate=startDate,
endDate=endDate,
freq=freq,
benchmark=benchmark,
logger=logger)
elif dataSource == DataSource.YAHOO:
dataHandler = YaHooDataProvider(symbolList=symbolList,
startDate=startDate,
endDate=endDate,
logger=logger)
backtest = Backtest(initialCapital,
0.0,
dataHandler,
SimulatedExecutionHandler,
Portfolio,
userStrategy,
logger,
benchmark,
refreshRate,
plot=plot,
portfolioType=portfolioType,
strategyParameters=strategyParameters)
equityCurve, \
orderBook, \
filledBook, \
perf_metric, \
perf_df, \
rollingRisk, \
aggregated_positions, \
transactions, \
turnover_rate, \
info_view \
= backtest.simulateTrading()
# save to a excel file
if saveFile:
if os.path.isdir('performance/'):
shutil.rmtree('performance/')
os.mkdir('performance/')
logger.info("Strategy performance is now saving to local files...")
perf_metric.to_csv('performance/perf_metrics.csv', float_format='%.4f')
perf_df.to_csv('performance/perf_series.csv', float_format='%.4f')
if benchmark is not None:
rollingRisk.to_csv('performance/rollingRisk.csv',
float_format='%.4f')
equityCurve.to_csv('performance/equity_curve.csv', float_format='%.4f')
orderBook.to_csv('performance/order_book.csv', float_format='%.4f')
filledBook.to_csv('performance/filled_book.csv', float_format='%.4f')
aggregated_positions.to_csv('performance/aggregated_positions.csv',
float_format='%.4f')
turnover_rate.to_csv('performance/turnover_rate.csv',
float_format='%.4f')
transactions.to_csv('performance/transactions.csv',
float_format='%.4f')
info_view.to_csv('performance/strategy_info.csv', float_format='%.4f')
logger.info("Performance saving is finished!")
return {
'equity_curve': equityCurve,
'order_book': orderBook,
'filled_book': filledBook,
'perf_metric': perf_metric,
'perf_series': perf_df,
'aggregated_positions': aggregated_positions,
'transactions': transactions,
'turnover_rate': turnover_rate,
'user_info': info_view
}
def strategyRunner(
userStrategy,
initialCapital=100000,
symbolList=["600000.XSHG"],
startDate=dt.datetime(2015, 9, 1),
endDate=dt.datetime(2015, 9, 15),
dataSource=DataSource.DXDataCenter,
benchmark=None,
refreshRate=1,
saveFile=False,
plot=False,
logLevel="info",
portfolioType=PortfolioType.CashManageable,
**kwargs
):
logger = CustomLogger(logLevel)
if dataSource == DataSource.CSV:
dataHandler = HistoricalCSVDataHandler(csvDir=kwargs["csvDir"], symbolList=symbolList, logger=logger)
elif dataSource == DataSource.DataYes:
try:
token = kwargs["token"]
except KeyError:
token = None
dataHandler = DataYesMarketDataHandler(
token=token, symbolList=symbolList, startDate=startDate, endDate=endDate, benchmark=benchmark, logger=logger
)
elif dataSource == DataSource.DXDataCenter:
dataHandler = DXDataCenter(
symbolList=symbolList,
startDate=startDate,
endDate=endDate,
freq=kwargs["freq"],
benchmark=benchmark,
logger=logger,
)
elif dataSource == DataSource.YAHOO:
dataHandler = YaHooDataProvider(symbolList=symbolList, startDate=startDate, endDate=endDate, logger=logger)
backtest = Backtest(
initialCapital,
0.0,
dataHandler,
SimulatedExecutionHandler,
Portfolio,
userStrategy,
logger,
benchmark,
refreshRate,
plot=plot,
portfolioType=portfolioType,
)
equityCurve, orderBook, filledBook, perf_metric, perf_df, rollingRisk, aggregated_positions, transactions, turnover_rate, info_view = (
backtest.simulateTrading()
)
# save to a excel file
if saveFile:
if os.path.isdir("performance/"):
shutil.rmtree("performance/")
os.mkdir("performance/")
logger.info("Strategy performance is now saving to local files...")
perf_metric.to_csv("performance/perf_metrics.csv", float_format="%.4f")
perf_df.to_csv("performance/perf_series.csv", float_format="%.4f")
if benchmark is not None:
rollingRisk.to_csv("performance/rollingRisk.csv", float_format="%.4f")
equityCurve.to_csv("performance/equity_curve.csv", float_format="%.4f")
orderBook.to_csv("performance/order_book.csv", float_format="%.4f")
filledBook.to_csv("performance/filled_book.csv", float_format="%.4f")
aggregated_positions.to_csv("performance/aggregated_positions.csv", float_format="%.4f")
turnover_rate.to_csv("performance/turnover_rate.csv", float_format="%.4f")
transactions.to_csv("performance/transactions.csv", float_format="%.4f")
info_view.to_csv("performance/strategy_info.csv", float_format="%.4f")
logger.info("Performance saving is finished!")
return {
"equity_curve": equityCurve,
"order_book": orderBook,
"filled_book": filledBook,
"perf_metric": perf_metric,
"perf_series": perf_df,
"aggregated_positions": aggregated_positions,
"transactions": transactions,
"turnover_rate": turnover_rate,
"user_info": info_view,
}
def strategyRunner(userStrategy,
strategyParameters=(),
initialCapital=100000,
symbolList=['600000.XSHG'],
startDate=dt.datetime(2015, 9, 1),
endDate=dt.datetime(2015, 9, 15),
benchmark=None,
refreshRate=1,
saveFile=False,
plot=True,
logLevel='critical',
portfolioType=PortfolioType.FullNotional,
**kwargs):
u"""
回测用运行器
:param userStrategy: 用户自定义策略
:param strategyParameters: 用户自定义策略所用参数，默认为None
:param initialCapital: 初始资金，默认为100000
:param symbolList: 证券代码，默认为600000.xshg
:param startDate: 回测起始日，默认为2015年9月1日
:param endDate: 回测结束日，默认为2015年9月15日
:param dataSource: 行情数据源，默认为DataSource.DXDataCenter
:param benchmark: 比较基准，比如指数。默认为None
:param refreshRate: bar的调用频率，默认为0
:param saveFile: 是否保存表现的数据为本地文件，默认为False
:param plot: 是否绘制表现图，默认为True
:param logLevel: 日志的等级
:param portfolioType: 收益计算的方法
:param kwargs: 其他需要的关键字参数
:return: dict
"""
logger = CustomLogger(logLevel)
try:
dataSource = kwargs['dataSource']
Settings.set_source(dataSource)
except KeyError:
dataSource = Settings.data_source
if dataSource == DataSource.CSV:
dataHandler = HistoricalCSVDataHandler(csvDir=kwargs['csvDir'],
symbolList=symbolList,
logger=logger)
elif dataSource == DataSource.DataYes:
try:
token = kwargs['token']
except KeyError:
token = None
dataHandler = DataYesMarketDataHandler(token=token,
symbolList=symbolList,
startDate=startDate,
endDate=endDate,
benchmark=benchmark,
logger=logger)
elif dataSource == DataSource.DXDataCenter:
try:
freq = kwargs['freq']
except KeyError:
freq = 0
logger.info("No `freq` keyword arguments found. using default value as freq=0")
dataHandler = DXDataCenter(symbolList=symbolList,
startDate=startDate,
endDate=endDate,
freq=freq,
benchmark=benchmark,
logger=logger)
elif dataSource == DataSource.YAHOO:
dataHandler = YaHooDataProvider(symbolList=symbolList,
startDate=startDate,
endDate=endDate,
logger=logger)
backtest = Backtest(initialCapital,
0.0,
dataHandler,
SimulatedExecutionHandler,
Portfolio,
userStrategy,
logger,
benchmark,
refreshRate,
plot=plot,
portfolioType=portfolioType,
strategyParameters=strategyParameters)
equityCurve, \
orderBook, \
filledBook, \
perf_metric, \
perf_df, \
rollingRisk, \
aggregated_positions, \
transactions, \
turnover_rate, \
info_view \
= backtest.simulateTrading()
# save to a excel file
if saveFile:
if os.path.isdir('performance/'):
shutil.rmtree('performance/')
os.mkdir('performance/')
logger.info("Strategy performance is now saving to local files...")
perf_metric.to_csv('performance/perf_metrics.csv', float_format='%.4f')
perf_df.to_csv('performance/perf_series.csv', float_format='%.4f')
if benchmark is not None:
rollingRisk.to_csv('performance/rollingRisk.csv', float_format='%.4f')
equityCurve.to_csv('performance/equity_curve.csv', float_format='%.4f')
orderBook.to_csv('performance/order_book.csv', float_format='%.4f')
filledBook.to_csv('performance/filled_book.csv', float_format='%.4f')
aggregated_positions.to_csv('performance/aggregated_positions.csv', float_format='%.4f')
turnover_rate.to_csv('performance/turnover_rate.csv', float_format='%.4f')
transactions.to_csv('performance/transactions.csv', float_format='%.4f')
info_view.to_csv('performance/strategy_info.csv', float_format='%.4f')
logger.info("Performance saving is finished!")
return {'equity_curve': equityCurve,
'order_book': orderBook,
'filled_book': filledBook,
'perf_metric': perf_metric,
'perf_series': perf_df,
'aggregated_positions': aggregated_positions,
'transactions': transactions,
'turnover_rate': turnover_rate,
'user_info': info_view}
def strategyRunner(userStrategy,
initialCapital=100000,
symbolList=['600000.XSHG'],
startDate=dt.datetime(2015, 9, 1),
endDate=dt.datetime(2015, 9, 15),
dataSource=DataSource.DXDataCenter,
benchmark=None,
refreshRate=1,
saveFile=False,
plot=True,
logLevel='critical',
portfolioType=PortfolioType.FullNotional,
**kwargs):
logger = CustomLogger(logLevel)
if dataSource == DataSource.CSV:
dataHandler = HistoricalCSVDataHandler(csvDir=kwargs['csvDir'],
symbolList=symbolList,
logger=logger)
elif dataSource == DataSource.DataYes:
try:
token = kwargs['token']
except KeyError:
token = None
dataHandler = DataYesMarketDataHandler(token=token,
symbolList=symbolList,
startDate=startDate,
endDate=endDate,
benchmark=benchmark,
logger=logger)
elif dataSource == DataSource.DXDataCenter:
try:
freq = kwargs['freq']
except KeyError:
freq = 0
dataHandler = DXDataCenter(symbolList=symbolList,
startDate=startDate,
endDate=endDate,
freq=freq,
benchmark=benchmark,
logger=logger)
elif dataSource == DataSource.YAHOO:
dataHandler = YaHooDataProvider(symbolList=symbolList,
startDate=startDate,
endDate=endDate,
logger=logger)
backtest = Backtest(initialCapital,
0.0,
dataHandler,
SimulatedExecutionHandler,
Portfolio,
userStrategy,
logger,
benchmark,
refreshRate,
plot=plot,
portfolioType=portfolioType)
equityCurve, \
orderBook, \
filledBook, \
perf_metric, \
perf_df, \
rollingRisk, \
aggregated_positions, \
transactions, \
turnover_rate, \
info_view \
= backtest.simulateTrading()
# save to a excel file
if saveFile:
if os.path.isdir('performance/'):
shutil.rmtree('performance/')
os.mkdir('performance/')
logger.info("Strategy performance is now saving to local files...")
perf_metric.to_csv('performance/perf_metrics.csv', float_format='%.4f')
perf_df.to_csv('performance/perf_series.csv', float_format='%.4f')
if benchmark is not None:
rollingRisk.to_csv('performance/rollingRisk.csv', float_format='%.4f')
equityCurve.to_csv('performance/equity_curve.csv', float_format='%.4f')
orderBook.to_csv('performance/order_book.csv', float_format='%.4f')
filledBook.to_csv('performance/filled_book.csv', float_format='%.4f')
aggregated_positions.to_csv('performance/aggregated_positions.csv', float_format='%.4f')
turnover_rate.to_csv('performance/turnover_rate.csv', float_format='%.4f')
transactions.to_csv('performance/transactions.csv', float_format='%.4f')
info_view.to_csv('performance/strategy_info.csv', float_format='%.4f')
logger.info("Performance saving is finished!")
return {'equity_curve': equityCurve,
'order_book': orderBook,
'filled_book': filledBook,
'perf_metric': perf_metric,
'perf_series': perf_df,
'aggregated_positions': aggregated_positions,
'transactions': transactions,
'turnover_rate': turnover_rate,
'user_info': info_view}
