VNPY 批量策略回測(cè)和統(tǒng)計(jì)結(jié)果的excel輸出是怎樣的

小編今天帶大家了解VNPY 批量策略回測(cè)和統(tǒng)計(jì)結(jié)果的excel輸出是怎樣的，文中知識(shí)點(diǎn)介紹的非常詳細(xì)。覺得有幫助的朋友可以跟著小編一起瀏覽文章的內(nèi)容，希望能夠幫助更多想解決這個(gè)問題的朋友找到問題的答案，下面跟著小編一起深入學(xué)習(xí)“VNPY 批量策略回測(cè)和統(tǒng)計(jì)結(jié)果的excel輸出是怎樣的”的知識(shí)吧。

做VNPY這段時(shí)間，發(fā)現(xiàn)主要就是回測(cè)，和策略優(yōu)化；然后就是有批量測(cè)試和參數(shù)集合效果導(dǎo)出excel分析要求。還有就是一個(gè)策略對(duì)不同品種效果驗(yàn)證。

然后想想，就自己寫了一個(gè)BatchBackTest 類，其實(shí)很簡(jiǎn)答，就是輸入品種隊(duì)列，和策略隊(duì)列，然后就會(huì)循環(huán)跑出結(jié)果，輸出到指定路徑excel。 還有就是策略是元組，是由一個(gè)策略和對(duì)應(yīng)參數(shù)組合。

代碼如下。使用方法就是在 vnpy/example/CtaBacktesting 路徑下面新疆一個(gè)py，放入下面代碼就可以

• # encoding: UTF-8
  

• from vnpy.trader.app.ctaStrategy.ctaBacktesting import BacktestingEngine, MINUTE_DB_NAME
  

• import pandas as pd
  

• from vnpy.trader.app.ctaStrategy.strategy.strategyAtrRsi import AtrRsiStrategy
  

• from vnpy.trader.app.ctaStrategy.strategy.strategyBollChannel import BollChannelStrategy
  

• from vnpy.trader.app.ctaStrategy.strategy.strategyDoubleMa import DoubleMaStrategy
  

• from vnpy.trader.app.ctaStrategy.strategy.strategyDualThrust import DualThrustStrategy
  

• from vnpy.trader.app.ctaStrategy.strategy.strategyKingKeltner import KkStrategy
  

• from vnpy.trader.app.ctaStrategy.strategy.strategyMultiSignal import MultiSignalStrategy
  

• from vnpy.trader.app.ctaStrategy.strategy.strategyMultiTimeframe import MultiTimeframeStrategy
  

• 
  

• class BatchBackTest(object):
  

•     def __init__(self):
  

•         ""
  

•     def calculateBacktesting(self,symbollist,strategylist):
  

•         #填入品種隊(duì)列和策略隊(duì)列，返回結(jié)果resultlist, 為了輸出方便檢索，加入品種名稱，策略名稱和策略參數(shù)
  

•         resultlist = []
  

•         for symbol in symbollist:
  

•             for strategy in strategylist:
  

•                 result = self.runBacktesting(symbol,strategy)
  

•                 #加入品種名稱，策略名稱和策略參數(shù)
  

•                 result["Symbolname"] = str(symbol["vtSymbol"])
  

•                 result["strategyname"] = str(strategy[0])
  

•                 result["strategysetting"] = str(strategy[1])
  

•                 resultlist.append(result)
  

•         return resultlist
  

• 
  

• 
  

•     def runBacktesting(self, symbol, strategy ):
  

•         #寫入測(cè)試品種和參數(shù), 返回回測(cè)數(shù)據(jù)集包含回測(cè)結(jié)果
  

• 
  

•         # 在引擎中創(chuàng)建策略對(duì)象
  

•         # 創(chuàng)建回測(cè)引擎
  

•         engine = BacktestingEngine()
  

•         # 設(shè)置引擎的回測(cè)模式為K線
  

•         engine.setBacktestingMode(engine.BAR_MODE)
  

•         # 設(shè)置回測(cè)用的數(shù)據(jù)起始日期
  

•         engine.setStartDate(symbol["StartDate"])
  

•         engine.setSlippage(symbol["Slippage"]) # 1跳
  

•         engine.setRate(symbol["Rate"]) # 傭金大小
  

•         engine.setSize(symbol["Size"]) # 合約大小
  

•         engine.setPriceTick(symbol["Slippage"]) # 最小價(jià)格變動(dòng)
  

•         engine.setCapital(symbol["Capital"])
  

• 
  

•         # 設(shè)置使用的歷史數(shù)據(jù)庫
  

•         engine.setDatabase(MINUTE_DB_NAME, symbol["vtSymbol"])
  

•         #設(shè)置策略，策略元組中第一個(gè)是策略，第二個(gè)參數(shù)
  

•         engine.initStrategy(strategy[0], strategy[1])
  

•         engine.runBacktesting()
  

•         df = engine.calculateDailyResult()
  

•         result = []
  

•         dfp,result = engine.calculateDailyStatistics(df)
  

•         engine.output(u'輸出統(tǒng)計(jì)數(shù)據(jù)')
  

•         # engine.showDailyResult(dfp, result)
  

•         return result
  

• 
  

•     def toExcel(self, resultlist, path = "C:\data\datframe.xlsx"):
  

•         #按照輸入統(tǒng)計(jì)數(shù)據(jù)隊(duì)列和路徑，輸出excel，這里不提供新增模式，如果想，可以改
  

•         #dft.to_csv(path,index=False,header=True, mode = 'a')
  

•         summayKey = resultlist[0].keys()
  

•         # summayValue = result.values()
  

• 
  

•         dft = pd.DataFrame(columns=summayKey)
  

•         for result in resultlist:
  

•             new = pd.DataFrame(result, index=["0"])
  

•             dft = dft.append(new,ignore_index=True)
  

•         dft.to_excel(path,index=False,header=True)
  

•         print "回測(cè)統(tǒng)計(jì)結(jié)果輸出到" + path
  

• 
  

• 
  

• if __name__ == "__main__":
  

•     #創(chuàng)建品種隊(duì)列，這里可以用json導(dǎo)入，為了方便使用直接寫了。
  

•     symbollist = [{
  

•         "vtSymbol": 'm1809',
  

•         "StartDate": "20180101",
  

•         "Slippage": 1,
  

•         "Size": 10,
  

•         "Rate": 2 / 10000,
  

•         "Capital": 10000
  

•             },
  

•         {
  

•             "vtSymbol": 'rb0000',
  

•             "StartDate": "20180101",
  

•             "Slippage": 1,
  

•             "Size": 10,
  

•             "Rate": 2 / 10000,
  

•             "Capital": 10000
  

•         }]
  

•     #這里定義策略，策略參數(shù)先為空；策略加參數(shù)是一個(gè)元組
  

•     setting = {}
  

•     Strategylist = [(AtrRsiStrategy, setting),
  

•                     (BollChannelStrategy, setting),
  

•                     (DoubleMaStrategy, setting),
  

•                     (DualThrustStrategy, setting),
  

•                     (KkStrategy, setting),
  

•                     (MultiSignalStrategy, setting),
  

•                     (MultiTimeframeStrategy, setting)]
  

•     # 這里是同一個(gè)策略，不同參數(shù)的情況，當(dāng)然可以有多個(gè)策略和多個(gè)參數(shù)組合
  

•     Strategylist2 = []
  

•     # 策略list
  

•     settinglist =[
  

•         {'kdlimit': 40, 'barmins': 9, 'cciWindow': 22},
  

•         {'kdlimit': 30, 'barmins': 13, 'cciWindow': 20}]
  

•     # 合并一個(gè)元組
  

•     if settinglist != []:
  

•         for para1 in settinglist:
  

•             Strategylist2.append((BollChannelStrategy, para1))
  

• 
  

•     NT = BatchBackTest()
  

•     resultlist = NT.calculateBacktesting(symbollist,Strategylist)
  

•     #定義路徑
  

•     path = "C:\Project\BackTestResult.xlsx"
  

•     NT.toExcel(resultlist,path)

感謝大家的閱讀，以上就是“VNPY 批量策略回測(cè)和統(tǒng)計(jì)結(jié)果的excel輸出是怎樣的”的全部?jī)?nèi)容了，學(xué)會(huì)的朋友趕緊操作起來吧。相信億速云小編一定會(huì)給大家?guī)砀鼉?yōu)質(zhì)的文章。謝謝大家對(duì)億速云網(wǎng)站的支持！