增加CTA模块下tools中的multiTimeFrame扩展功能

vn.trader/ctaAlgo/tools/README.md

@@ -5,3 +5,6 @@
 * 贡献者：李来佳
 * WeChat/QQ: 28888502
 
+### multiTimeFrame
+* 简介：基于CTA模块扩展了回测和交易功能，允许策略中引用辅助品种信息（其他时间框架、其他合约），同时提供了一个突破策略的例子
+* 贡献者：周正舟

vn.trader/ctaAlgo/tools/multiTimeFrame/ctaBacktestMultiTF.py

@@ -0,0 +1,206 @@
+# encoding: UTF-8
+
+'''
+这个文件加入在CTA回测引擎的基础上加入了辅助品种信息, 保持接口的一致, 可以在原CTA引擎上执行的代码,
+也可以在这个引擎上执行
+This file add multi Time Frame functionalities to CTA backtesting engine, the APIs are the
+same as CTA engine. Real trading code can be directly used for backtesting.
+'''
+
+from __future__ import division
+from vtFunction import loadMongoSetting
+
+from ctaBacktesting import *
+
+class BacktestEngineMultiTF(BacktestingEngine):
+
+    def __init__(self):
+        """Constructor"""
+        super(BacktestEngineMultiTF, self).__init__()
+
+        self.info_symbols   = []        # List, 输入辅助品种的2值tuple, 左边为数据库名, 右边为collection名
+        self.InfoCursor     = {}        # Dict, 放置回测用辅助品种数据库
+        self.initInfoCursor = {}        # Dict, 放置初始化用辅助品种数据库
+        self.infobar        = {}        # Dict, 放置辅助品种最新一个K线数据
+        self.MultiOn        = False     # Boolean, 判断是否传入了辅助品种
+
+    # ----------------------------------------------------------------------
+    def setDatabase(self, dbName, symbol, **kwargs):
+        """set database that provide historical data"""
+
+        self.dbName = dbName
+
+        # Set executed symbol and information symbols
+        self.symbol = symbol
+        if "info_symbol" in kwargs:
+            self.info_symbols = kwargs["info_symbol"]
+
+            # Turn on MultiTF switch
+            if len(self.info_symbols) > 0:
+                self.MultiOn = True
+
+    # ----------------------------------------------------------------------
+    def loadInitData(self, collection, **kwargs):
+        """Load initializing data"""
+        # 载入初始化需要用的数据
+        # Load initialised data
+
+        # $gte means "greater and equal to"
+        # $lt means "less than"
+        flt = {'datetime': {'$gte': self.dataStartDate,
+                            '$lt': self.strategyStartDate}}
+        self.initCursor = collection.find(flt)
+
+        # 初始化辅助品种数据
+        # Initializing information data
+        if "inf" in kwargs:
+            for name in kwargs["inf"]:
+                DB = kwargs["inf"][name]
+                self.initInfoCursor[name] = DB.find(flt)
+
+        # 将数据从查询指针中读取出，并生成列表
+        # Read data from cursor, generate a list
+        self.initData = []
+
+        for d in self.initCursor:
+            data = self.dataClass()
+            data.__dict__ = d
+            self.initData.append(data)
+
+    # ----------------------------------------------------------------------
+    def loadHistoryData(self):
+        """载入历史数据"""
+        """load historical data"""
+
+        host, port = loadMongoSetting()
+
+        self.dbClient = pymongo.MongoClient(host, port)
+        collection = self.dbClient[self.dbName][self.symbol]
+
+        # Load historical data of information symbols, construct a dictionary of Database
+        # Values of dictionary are mongo.Client.
+        info_collection = {}
+        if self.MultiOn is True:
+            for DBname, symbol in self.info_symbols:
+                info_collection[DBname + " " + symbol] = self.dbClient[DBname][symbol]
+
+        self.output("Start loading historical data")
+
+        # 首先根据回测模式，确认要使用的数据类
+        # Choose data type based on backtest mode
+        if self.mode == self.BAR_MODE:
+            self.dataClass = CtaBarData
+            self.func      = self.newBar
+        else:
+            self.dataClass = CtaTickData
+            self.func = self.newTick
+
+        # Load initializing data
+        self.loadInitData(collection, inf=info_collection)
+
+        # 载入回测数据
+        # Load backtest data (exclude initializing data)
+        if not self.dataEndDate:
+            # If "End Date" is not set, retreat data up to today
+            flt = {'datetime': {'$gte': self.strategyStartDate}}
+        else:
+            flt = {'datetime': {'$gte': self.strategyStartDate,
+                                '$lte': self.dataEndDate}}
+        self.dbCursor = collection.find(flt)
+
+        if self.MultiOn is True:
+            for db in info_collection:
+                self.InfoCursor[db] = info_collection[db].find(flt)
+            self.output(
+                "Data loading completed, data volumn: %s" % (self.initCursor.count() + self.dbCursor.count() + \
+                                                             sum([i.count() for i in self.InfoCursor.values()])))
+        else:
+            self.output("Data loading completed, data volumn: %s" % (self.initCursor.count() + self.dbCursor.count()))
+
+    # ----------------------------------------------------------------------
+    def runBacktesting(self):
+        """运行回测"""
+        """Run backtesting"""
+
+        # 载入历史数据
+        # Load historical data
+        self.loadHistoryData()
+
+        self.output("Start backtesing!")
+
+        self.strategy.inited = True
+        self.strategy.onInit()
+        self.output("Strategy initialsing complete")
+
+        self.strategy.trading = True
+        self.strategy.onStart()
+        self.output("Strategy started")
+
+        self.output("Processing historical data...")
+
+        dataClass = self.dataClass
+        func = self.func
+        for d in self.dbCursor:
+            data = dataClass()
+            data.__dict__ = d
+            func(data)
+
+        self.output("No more historical data")
+
+    # ----------------------------------------------------------------------
+    def checkInformationData(self):
+        """Update information symbols' data"""
+
+        # If infobar is empty, which means it is the first time calling this method
+        if self.infobar == {}:
+            for info_symbol in self.InfoCursor:
+                try:
+                    self.infobar[info_symbol] = next(self.InfoCursor[info_symbol])
+                except StopIteration:
+                    print "Data of information symbols is empty! Input must be a list, not str."
+                    raise
+
+        temp = {}
+        for info_symbol in self.infobar:
+
+            data = self.infobar[info_symbol]
+
+            # Update data only when Time Stamp is matched
+            if data['datetime'] <= self.dt:
+                try:
+                    temp[info_symbol] = CtaBarData()
+                    temp[info_symbol].__dict__ = data
+                    self.infobar[info_symbol] = next(self.InfoCursor[info_symbol])
+                except StopIteration:
+                    self.output("No more data in information database.")
+            else:
+                temp[info_symbol] = None
+
+        return temp
+
+    # ----------------------------------------------------------------------
+    def newBar(self, bar):
+        """新的K线"""
+        """new ohlc Bar"""
+        self.bar = bar
+        self.dt = bar.datetime
+        self.updatePosition()  # Update total position value based on new Bar
+        self.crossLimitOrder()  # 先撮合限价单
+        self.crossStopOrder()  # 再撮合停止单
+        if self.MultiOn is True:
+            self.strategy.onBar(bar, infobar=self.checkInformationData())  # 推送K线到策略中
+        else:
+            self.strategy.onBar(bar)  # 推送K线到策略中
+
+    # ----------------------------------------------------------------------
+    def newTick(self, tick):
+        """新的Tick"""
+        """new Tick"""
+        self.tick = tick
+        self.dt = tick.datetime
+        self.crossLimitOrder()
+        self.crossStopOrder()
+        self.strategy.onTick(tick)
+
+########################################################################
+

vn.trader/ctaAlgo/tools/multiTimeFrame/ctaStrategyMultiTF.py

@@ -0,0 +1,410 @@
+# encoding: UTF-8
+"""
+This file tweaks ctaTemplate Module to suit multi-TimeFrame strategies.
+"""
+
+from strategyAtrRsi import *
+from ctaBase import *
+from ctaTemplate import CtaTemplate
+
+########################################################################
+class TC11(CtaTemplate):
+
+    # Strategy name and author
+    className = "TC11"
+    author = "Zenacon"
+
+    # Set MongoDB DataBase
+    barDbName = "TestData"
+
+    # Strategy parameters
+    pGeneric_prd = 21
+    pGeneric_on = True
+
+    pATRprd_F = 13
+    pATRprd_M = 21
+    pATRprd_S = 63
+
+    pBOSSplus_prd = 98
+    pBOSSminus_prd = 22
+
+    if pGeneric_on == 0:
+        pRSIprd = 20
+        pBBprd = 10
+        pBB_ATRprd = 15
+        pATRprd = 21
+        pDMIprd = 21
+    else:
+        pRSIprd =       \
+        pBBprd =        \
+        pBB_ATRprd =    \
+        pATRprd =       \
+        pDMIprd = pGeneric_prd
+
+    pBOSS_Mult = 1.75
+
+    # Strategy variables
+    vOBO_initialpoint = EMPTY_FLOAT
+    vOBO_Stretch = EMPTY_FLOAT
+    vOBO_level_L = EMPTY_FLOAT
+    vOBO_level_S = EMPTY_FLOAT
+
+    # parameters' list, record names of parameters
+    paramList = ['name',
+                 'className',
+                 'author',
+                 'vtSymbol']
+
+    # variables' list, record names of variables
+    varList = ['inited',
+               'trading',
+               'pos']
+
+    def __init__(self, ctaEngine, setting):
+        """Constructor"""
+        super(TC11, self).__init__(ctaEngine, setting)
+
+    # ----------------------------------------------------------------------
+    def onBar(self, bar, **kwargs):
+        """收到Bar推送（必须由用户继承实现）"""
+        # 撤销之前发出的尚未成交的委托（包括限价单和停止单）
+        for orderID in self.orderList:
+            self.cancelOrder(orderID)
+        self.orderList = []
+
+        # Record new information bar
+        if "infobar" in kwargs:
+            for i in kwargs["infobar"]:
+                if kwargs["infobar"][i] is None:
+                    pass
+                else:
+                    # print kwargs["infobar"][i]["close"]
+                    self.closeArray[0:self.bufferSize - 1] = self.closeArray[1:self.bufferSize]
+                    self.highArray[0:self.bufferSize - 1] = self.highArray[1:self.bufferSize]
+                    self.lowArray[0:self.bufferSize - 1] = self.lowArray[1:self.bufferSize]
+
+                    self.closeArray[-1] = bar.close
+                    self.highArray[-1] = bar.high
+                    self.lowArray[-1] = bar.low
+
+        """
+        Record new bar
+        """
+        self.closeArray[0:self.bufferSize - 1] = self.closeArray[1:self.bufferSize]
+        self.highArray[0:self.bufferSize - 1] = self.highArray[1:self.bufferSize]
+        self.lowArray[0:self.bufferSize - 1] = self.lowArray[1:self.bufferSize]
+
+        self.closeArray[-1] = bar.close
+        self.highArray[-1] = bar.high
+        self.lowArray[-1] = bar.low
+
+        self.bufferCount += 1
+        if self.bufferCount < self.bufferSize:
+            return
+
+        """
+        Calculate Indicators
+        """
+
+        vOBO_initialpoint = self.dataHTF_filled['Open']
+        vOBO_Stretch = self.vATR['htf'].m * self.pBOSS_Mult
+
+        self.atrValue = talib.ATR(self.highArray,
+                                  self.lowArray,
+                                  self.closeArray,
+                                  self.atrLength)[-1]
+        self.atrArray[0:self.bufferSize - 1] = self.atrArray[1:self.bufferSize]
+        self.atrArray[-1] = self.atrValue
+
+        self.atrCount += 1
+        if self.atrCount < self.bufferSize:
+            return
+
+        self.atrMa = talib.MA(self.atrArray,
+                              self.atrMaLength)[-1]
+        self.rsiValue = talib.RSI(self.closeArray,
+                                  self.rsiLength)[-1]
+
+        # 判断是否要进行交易
+
+        # 当前无仓位
+        if self.pos == 0:
+            self.intraTradeHigh = bar.high
+            self.intraTradeLow = bar.low
+
+            # ATR数值上穿其移动平均线，说明行情短期内波动加大
+            # 即处于趋势的概率较大，适合CTA开仓
+            if self.atrValue > self.atrMa:
+                # 使用RSI指标的趋势行情时，会在超买超卖区钝化特征，作为开仓信号
+                if self.rsiValue > self.rsiBuy:
+                    # 这里为了保证成交，选择超价5个整指数点下单
+                    self.buy(bar.close + 5, 1)
+
+                elif self.rsiValue < self.rsiSell:
+                    self.short(bar.close - 5, 1)
+
+        # 持有多头仓位
+        elif self.pos > 0:
+            # 计算多头持有期内的最高价，以及重置最低价
+            self.intraTradeHigh = max(self.intraTradeHigh, bar.high)
+            self.intraTradeLow = bar.low
+            # 计算多头移动止损
+            longStop = self.intraTradeHigh * (1 - self.trailingPercent / 100)
+            # 发出本地止损委托，并且把委托号记录下来，用于后续撤单
+            orderID = self.sell(longStop, 1, stop=True)
+            self.orderList.append(orderID)
+
+        # 持有空头仓位
+        elif self.pos < 0:
+            self.intraTradeLow = min(self.intraTradeLow, bar.low)
+            self.intraTradeHigh = bar.high
+
+            shortStop = self.intraTradeLow * (1 + self.trailingPercent / 100)
+            orderID = self.cover(shortStop, 1, stop=True)
+            self.orderList.append(orderID)
+
+        # 发出状态更新事件
+        self.putEvent()
+
+########################################################################
+class Prototype(AtrRsiStrategy):
+
+    """
+    "infoArray" 字典是用来储存辅助品种信息的, 可以是同品种的不同分钟k线, 也可以是不同品种的价格。
+
+    调用的方法:
+    self.infoArray["数据库名 + 空格 + collection名"]["close"]
+    self.infoArray["数据库名 + 空格 + collection名"]["high"]
+    self.infoArray["数据库名 + 空格 + collection名"]["low"]
+    """
+    infoArray = {}
+    initInfobar = {}
+
+    def __int__(self):
+        super(Prototype, self).__int__()
+
+    # ----------------------------------------------------------------------
+    def onInit(self):
+        """初始化策略（必须由用户继承实现）"""
+        self.writeCtaLog(u'%s策略初始化' % self.name)
+
+        # 初始化RSI入场阈值
+        self.rsiBuy = 50 + self.rsiEntry
+        self.rsiSell = 50 - self.rsiEntry
+
+        # 载入历史数据，并采用回放计算的方式初始化策略数值
+        initData = self.loadBar(self.initDays)
+        for bar in initData:
+
+            # 推送新数据, 同时检查是否有information bar需要推送
+            # Update new bar, check whether the Time Stamp matching any information bar
+            ibar = self.checkInfoBar(bar)
+            self.onBar(bar, infobar=ibar)
+
+        self.putEvent()
+
+    # ----------------------------------------------------------------------
+    def checkInfoBar(self, bar):
+        """在初始化时, 检查辅助品种数据的推送(初始化结束后, 回测时不会调用)"""
+
+        initInfoCursorDict = self.ctaEngine.initInfoCursor
+
+        # 如果"initInfobar"字典为空, 初始化字典, 插入第一个数据
+        # If dictionary "initInfobar" is empty, insert first data record
+        if self.initInfobar == {}:
+            for info_symbol in initInfoCursorDict:
+                try:
+                    self.initInfobar[info_symbol] = next(initInfoCursorDict[info_symbol])
+                except StopIteration:
+                    print "Data of information symbols is empty! Input is a list, not str."
+                    raise
+
+        # 若有某一品种的 TimeStamp 和执行报价的 TimeStamp 匹配, 则将"initInfobar"中的数据推送,
+        # 然后更新该品种的数据
+        # If any symbol's TimeStamp is matched with execution symbol's TimeStamp, return data
+        # in "initInfobar", and update new data.
+        temp = {}
+        for info_symbol in self.initInfobar:
+
+            data = self.initInfobar[info_symbol]
+
+            # Update data only when Time Stamp is matched
+            if data['datetime'] <= bar.datetime:
+                try:
+                    temp[info_symbol] = CtaBarData()
+                    temp[info_symbol].__dict__ = data
+                    self.initInfobar[info_symbol] = next(initInfoCursorDict[info_symbol])
+                except StopIteration:
+                    self.ctaEngine.output("No more data for initializing %s." % (info_symbol,))
+            else:
+                temp[info_symbol] = None
+
+        return temp
+
+    # ----------------------------------------------------------------------
+    def updateInfoArray(self, infobar):
+        """收到Infomation Data, 更新辅助品种缓存字典"""
+
+        for name in infobar:
+
+            data = infobar[name]
+
+            # Construct empty array
+            if len(self.infoArray) < len(infobar) :
+                self.infoArray[name] = {
+                    "close": np.zeros(self.bufferSize),
+                    "high": np.zeros(self.bufferSize),
+                    "low": np.zeros(self.bufferSize)
+                }
+
+            if data is None:
+                pass
+
+            else:
+                self.infoArray[name]["close"][0:self.bufferSize - 1] = \
+                    self.infoArray[name]["close"][1:self.bufferSize]
+                self.infoArray[name]["high"][0:self.bufferSize - 1] = \
+                    self.infoArray[name]["high"][1:self.bufferSize]
+                self.infoArray[name]["low"][0:self.bufferSize - 1] = \
+                    self.infoArray[name]["low"][1:self.bufferSize]
+
+                self.infoArray[name]["close"][-1] = data.close
+                self.infoArray[name]["high"][-1] = data.high
+                self.infoArray[name]["low"][-1] = data.low
+
+    # ----------------------------------------------------------------------
+    def onBar(self, bar, **kwargs):
+        """收到Bar推送（必须由用户继承实现）"""
+        # 撤销之前发出的尚未成交的委托（包括限价单和停止单）
+        for orderID in self.orderList:
+            self.cancelOrder(orderID)
+        self.orderList = []
+
+        # Update infomation data
+        # "infobar"是由不同时间或不同品种的品种数据组成的字典, 如果和执行品种的 TimeStamp 不匹配,
+        # 则传入的是"None", 当time stamp和执行品种匹配时, 传入的是"Bar"
+        self.updateInfoArray(kwargs["infobar"])
+
+        # 保存K线数据
+        self.closeArray[0:self.bufferSize - 1] = self.closeArray[1:self.bufferSize]
+        self.highArray[0:self.bufferSize - 1] = self.highArray[1:self.bufferSize]
+        self.lowArray[0:self.bufferSize - 1] = self.lowArray[1:self.bufferSize]
+
+        self.closeArray[-1] = bar.close
+        self.highArray[-1] = bar.high
+        self.lowArray[-1] = bar.low
+
+        # 若读取的缓存数据不足, 不考虑交易
+        self.bufferCount += 1
+        if self.bufferCount < self.bufferSize:
+            return
+
+        # 计算指标数值
+
+        # 计算不同时间下的ATR数值
+
+        # Only trading when information bar changes
+        # 只有在30min或者1d K线更新后才可以交易
+        TradeOn = False
+        if any([i is not None for i in kwargs["infobar"].values()]):
+
+            TradeOn = True
+            self.scaledAtrValue1M = talib.ATR(self.highArray,
+                                       self.lowArray,
+                                       self.closeArray,
+                                       self.atrLength)[-1] * (25) ** (0.5)
+            self.atrValue30M = talib.abstract.ATR(self.infoArray["TestData @GC_30M"])[-1]
+            self.rsiValue = talib.abstract.RSI(self.infoArray["TestData @GC_30M"], self.rsiLength)[-1]
+
+        self.atrCount += 1
+        if self.atrCount < self.bufferSize:
+            return
+
+        # 判断是否要进行交易
+
+        # 当前无仓位
+        if (self.pos == 0 and TradeOn == True):
+            self.intraTradeHigh = bar.high
+            self.intraTradeLow = bar.low
+
+            # 1Min调整后ATR大于30MinATR
+            # 即处于趋势的概率较大，适合CTA开仓
+            if self.atrValue30M < self.scaledAtrValue1M:
+                # 使用RSI指标的趋势行情时，会在超买超卖区钝化特征，作为开仓信号
+                if self.rsiValue > self.rsiBuy:
+                    # 这里为了保证成交，选择超价5个整指数点下单
+                    self.buy(bar.close+5, 1)
+
+                elif self.rsiValue < self.rsiSell:
+                    self.short(bar.close-5, 1)
+
+                # 下单后, 在下一个30Min K线之前不交易
+                TradeOn = False
+
+        # 持有多头仓位
+        elif self.pos > 0:
+            # 计算多头持有期内的最高价，以及重置最低价
+            self.intraTradeHigh = max(self.intraTradeHigh, bar.high)
+            self.intraTradeLow = bar.low
+            # 计算多头移动止损
+            longStop = self.intraTradeHigh * (1 - self.trailingPercent / 100)
+            # 发出本地止损委托，并且把委托号记录下来，用于后续撤单
+            orderID = self.sell(longStop, 1, stop=True)
+            self.orderList.append(orderID)
+
+        # 持有空头仓位
+        elif self.pos < 0:
+            self.intraTradeLow = min(self.intraTradeLow, bar.low)
+            self.intraTradeHigh = bar.high
+
+            shortStop = self.intraTradeLow * (1 + self.trailingPercent / 100)
+            orderID = self.cover(shortStop, 1, stop=True)
+            self.orderList.append(orderID)
+
+        # 发出状态更新事件
+        self.putEvent()
+
+
+if __name__ == '__main__':
+    # 提供直接双击回测的功能
+    # 导入PyQt4的包是为了保证matplotlib使用PyQt4而不是PySide，防止初始化出错
+    from ctaBacktestMultiTF import *
+    from PyQt4 import QtCore, QtGui
+    import time
+
+    '''
+    创建回测引擎
+    设置引擎的回测模式为K线
+    设置回测用的数据起始日期
+    载入历史数据到引擎中
+    在引擎中创建策略对象
+
+    Create backtesting engine
+    Set backtest mode as "Bar"
+    Set "Start Date" of data range
+    Load historical data to engine
+    Create strategy instance in engine
+    '''
+    engine = BacktestEngineMultiTF()
+    engine.setBacktestingMode(engine.BAR_MODE)
+    engine.setStartDate('20100101')
+    engine.setDatabase("TestData", "@GC_1M", info_symbol=[("TestData","@GC_30M")])
+
+    # Set parameters for strategy
+    d = {'atrLength': 11}
+    engine.initStrategy(Prototype, d)
+
+    # 设置产品相关参数
+    engine.setSlippage(0.2)  # 股指1跳
+    engine.setCommission(0.3 / 10000)  # 万0.3
+    engine.setSize(300)  # 股指合约大小
+
+    # 开始跑回测
+    start = time.time()
+
+    engine.runBacktesting()
+
+    # 显示回测结果
+    engine.showBacktestingResult()
+
+    print 'Time consumed：%s' % (time.time() - start)

vn.trader/ctaAlgo/tools/multiTimeFrame/strategyBreakOut.py

@@ -0,0 +1,317 @@
+# encoding: UTF-8
+"""
+This file tweaks ctaTemplate Module to suit multi-TimeFrame strategies.
+"""
+
+from ctaBase import *
+from ctaTemplate import CtaTemplate
+import numpy as np
+
+########################################################################
+class BreakOut(CtaTemplate):
+
+    """
+    "infoArray" 字典是用来储存辅助品种信息的, 可以是同品种的不同分钟k线, 也可以是不同品种的价格。
+
+    调用的方法:
+    价格序列:
+    self.infoArray["数据库名 + 空格 + collection名"]["close"]
+    self.infoArray["数据库名 + 空格 + collection名"]["high"]
+    self.infoArray["数据库名 + 空格 + collection名"]["low"]
+
+    单个价格:
+    self.infoBar["数据库名 + 空格 + collection名"]
+    返回的值为一个ctaBarData 或 None
+    """
+
+    #----------------------------------------------------------------------
+    def __init__(self, ctaEngine, setting):
+        """日内突破交易策略, 出场方式非常多, 本文件使用指标出场"""
+
+        className = 'BreakOut'
+        author = 'Joe'
+        super(BreakOut, self).__init__(ctaEngine, setting)
+
+        # 设置辅助品种数据字典
+        self.infoArray = {}
+        self.initInfobar = {}
+        self.infoBar = {}
+
+        # 缓存数据量
+        self.bufferSize = 100
+        self.bufferCount = 0
+        self.initDays = 10
+
+        # 设置参数
+        self.pOBO_Mult = 0.5        # 计算突破点位
+        # self.pProtMult = 2          # 止损的ATR倍数
+        # self.pProfitMult = 2        # 止盈相对于止损的倍数
+        # self.SlTp_On = False        # 止损止盈功能
+        # self.EODTime = 15           # 设置日内平仓时间
+
+        self.vOBO_stretch = EMPTY_FLOAT
+        self.vOBO_initialpoint = EMPTY_FLOAT
+        self.vOBO_level_L = EMPTY_FLOAT
+        self.vOBO_level_S = EMPTY_FLOAT
+
+        self.orderList = []
+
+        # 参数列表，保存了参数的名称
+        paramList = ['name',
+                     'className',
+                     'author',
+                     'pOBO_Mult',
+                     'pProtMult',
+                     'pProfitMult',
+                     'SlTp_On',
+                     'EODTime']
+
+        # 变量列表，保存了变量的名称
+        varList = ['vOBO_stretch',
+                   'vOBO_initialpoint',
+                   'vOBO_level_L',
+                   'vOBO_level_S']
+
+        # ----------------------------------------------------------------------
+    def onInit(self):
+        """初始化策略（必须由用户继承实现）"""
+        self.writeCtaLog(u'%s策略初始化' % self.name)
+
+        # 载入历史数据，并采用回放计算的方式初始化策略数值
+        initData = self.loadBar(self.initDays)
+        for bar in initData:
+
+            # 推送新数据, 同时检查是否有information bar需要推送
+            # Update new bar, check whether the Time Stamp matching any information bar
+            ibar = self.checkInfoBar(bar)
+            self.onBar(bar, infobar=ibar)
+
+        self.putEvent()
+
+    #----------------------------------------------------------------------
+    def onStart(self):
+        """启动策略（必须由用户继承实现）"""
+        self.writeCtaLog(u'%s策略启动' %self.name)
+        self.putEvent()
+
+    #----------------------------------------------------------------------
+    def onStop(self):
+        """停止策略（必须由用户继承实现）"""
+        self.writeCtaLog(u'%s策略停止' %self.name)
+        self.putEvent()
+
+    # ----------------------------------------------------------------------
+    def checkInfoBar(self, bar):
+        """在初始化时, 检查辅助品种数据的推送(初始化结束后, 回测时不会调用)"""
+
+        initInfoCursorDict = self.ctaEngine.initInfoCursor
+
+        # 如果"initInfobar"字典为空, 初始化字典, 插入第一个数据
+        # If dictionary "initInfobar" is empty, insert first data record
+        if self.initInfobar == {}:
+            for info_symbol in initInfoCursorDict:
+                try:
+                    self.initInfobar[info_symbol] = next(initInfoCursorDict[info_symbol])
+                except StopIteration:
+                    print "Data of information symbols is empty! Input is a list, not str."
+                    raise
+
+        # 若有某一品种的 TimeStamp 和执行报价的 TimeStamp 匹配, 则将"initInfobar"中的数据推送,
+        # 然后更新该品种的数据
+        # If any symbol's TimeStamp is matched with execution symbol's TimeStamp, return data
+        # in "initInfobar", and update new data.
+        temp = {}
+        for info_symbol in self.initInfobar:
+
+            data = self.initInfobar[info_symbol]
+
+            # Update data only when Time Stamp is matched
+
+            if (data is not None) and (data['datetime'] <= bar.datetime):
+
+                try:
+                    temp[info_symbol] = CtaBarData()
+                    temp[info_symbol].__dict__ = data
+                    self.initInfobar[info_symbol] = next(initInfoCursorDict[info_symbol])
+                except StopIteration:
+                    self.initInfobar[info_symbol] = None
+                    self.ctaEngine.output("No more data for initializing %s." % (info_symbol,))
+            else:
+                temp[info_symbol] = None
+
+        return temp
+
+    # ----------------------------------------------------------------------
+    def updateInfoArray(self, infobar):
+        """收到Infomation Data, 更新辅助品种缓存字典"""
+
+        for name in infobar:
+
+            data = infobar[name]
+
+            # Construct empty array
+            if len(self.infoArray) < len(infobar) :
+                self.infoArray[name] = {
+                    "close": np.zeros(self.bufferSize),
+                    "high": np.zeros(self.bufferSize),
+                    "low": np.zeros(self.bufferSize),
+                    "open": np.zeros(self.bufferSize)
+                }
+
+            if data is None:
+                pass
+
+            else:
+                self.infoArray[name]["close"][0:self.bufferSize - 1] = \
+                    self.infoArray[name]["close"][1:self.bufferSize]
+                self.infoArray[name]["high"][0:self.bufferSize - 1] = \
+                    self.infoArray[name]["high"][1:self.bufferSize]
+                self.infoArray[name]["low"][0:self.bufferSize - 1] = \
+                    self.infoArray[name]["low"][1:self.bufferSize]
+                self.infoArray[name]["open"][0:self.bufferSize - 1] = \
+                    self.infoArray[name]["open"][1:self.bufferSize]
+
+                self.infoArray[name]["close"][-1] = data.close
+                self.infoArray[name]["high"][-1] = data.high
+                self.infoArray[name]["low"][-1] = data.low
+                self.infoArray[name]["open"][-1] = data.open
+
+    # ----------------------------------------------------------------------
+    def onBar(self, bar, **kwargs):
+        """收到Bar推送（必须由用户继承实现）"""
+
+        # Update infomation data
+        # "infobar"是由不同时间或不同品种的品种数据组成的字典, 如果和执行品种的 TimeStamp 不匹配,
+        # 则传入的是"None", 当time stamp和执行品种匹配时, 传入的是"Bar"
+        if "infobar" in kwargs:
+            self.infoBar = kwargs["infobar"]
+            self.updateInfoArray(kwargs["infobar"])
+
+        # 若读取的缓存数据不足, 不考虑交易
+        self.bufferCount += 1
+        if self.bufferCount < self.bufferSize:
+            return
+
+        # 计算指标数值
+        a = np.sum(self.infoArray["TestData @GC_1D"]["close"])
+        if a == 0.0:
+            return
+
+        # Only updating indicators when information bar changes
+        # 只有在30min或者1d K线更新后才更新指标
+        TradeOn = False
+        if any([i is not None for i in self.infoBar]):
+            TradeOn = True
+            self.vRange = self.infoArray["TestData @GC_1D"]["high"][-1] -\
+                          self.infoArray["TestData @GC_1D"]["low"][-1]
+            self.vOBO_stretch = self.vRange * self.pOBO_Mult
+            self.vOBO_initialpoint = self.infoArray["TestData @GC_1D"]["close"][-1]
+            self.vOBO_level_L = self.vOBO_initialpoint + self.vOBO_stretch
+            self.vOBO_level_S = self.vOBO_initialpoint - self.vOBO_stretch
+
+            self.atrValue30M = talib.abstract.ATR(self.infoArray["TestData @GC_30M"])[-1]
+
+        # 判断是否要进行交易
+
+        # 当前无仓位
+        if (self.pos == 0 and TradeOn == True):
+
+            # 撤销之前发出的尚未成交的委托（包括限价单和停止单）
+            for orderID in self.orderList:
+                self.cancelOrder(orderID)
+            self.orderList = []
+
+            # 若上一个30分钟K线的最高价大于OBO_level_L
+            # 且当前的价格大于OBO_level_L, 则买入
+            if self.infoArray["TestData @GC_30M"]["high"][-1] > self.vOBO_level_L:
+
+                if bar.close > self.vOBO_level_L:
+
+                    self.buy(bar.close + 0.5, 1)
+
+                    # 下单后, 在下一个30Min K线之前不交易
+                    TradeOn = False
+
+            # 若上一个30分钟K线的最高价低于OBO_level_S
+            # 且当前的价格小于OBO_level_S, 则卖出
+            elif self.infoArray["TestData @GC_30M"]["low"][-1] < self.vOBO_level_S:
+
+                if bar.close < self.vOBO_level_S:
+
+                    self.short(bar.close - 0.5, 1)
+
+                    # 下单后, 在下一个30Min K线之前不交易
+                    TradeOn = False
+
+        # 持有多头仓位
+        elif self.pos > 0:
+
+            # 当价格低于initialpoint水平, 出场
+            if bar.close < self.vOBO_initialpoint:
+                self.sell(bar.close - 0.5 , 1)
+
+        # 持有空头仓位
+        elif self.pos < 0:
+
+            # 当价格高于initialpoint水平, 出场
+            if bar.close > self.vOBO_initialpoint:
+                self.cover(bar.close + 0.5, 1)
+
+
+        # 发出状态更新事件
+        self.putEvent()
+
+    # ----------------------------------------------------------------------
+    def onOrder(self, order):
+        """收到委托变化推送（必须由用户继承实现）"""
+        pass
+
+    # ----------------------------------------------------------------------
+    def onTrade(self, trade):
+        pass
+
+
+if __name__ == '__main__':
+    # 提供直接双击回测的功能
+    # 导入PyQt4的包是为了保证matplotlib使用PyQt4而不是PySide，防止初始化出错
+    from ctaBacktestMultiTF import *
+    from PyQt4 import QtCore, QtGui
+    import time
+
+    '''
+    创建回测引擎
+    设置引擎的回测模式为K线
+    设置回测用的数据起始日期
+    载入历史数据到引擎中
+    在引擎中创建策略对象
+
+    Create backtesting engine
+    Set backtest mode as "Bar"
+    Set "Start Date" of data range
+    Load historical data to engine
+    Create strategy instance in engine
+    '''
+    engine = BacktestEngineMultiTF()
+    engine.setBacktestingMode(engine.BAR_MODE)
+    engine.setStartDate('20120101')
+    engine.setEndDate('20150101')
+    engine.setDatabase("TestData", "@GC_1M", info_symbol=[("TestData","@GC_30M"),
+                                                          ("TestData","@GC_1D")])
+
+    # Set parameters for strategy
+    engine.initStrategy(BreakOut, {})
+
+    # 设置产品相关参数
+    engine.setSlippage(0.2)  # 股指1跳
+    engine.setCommission(0.3 / 10000)  # 万0.3
+    engine.setSize(1)  # 股指合约大小
+
+    # 开始跑回测
+    start = time.time()
+
+    engine.runBacktesting()
+
+    # 显示回测结果
+    engine.showBacktestingResult()
+
+    print 'Time consumed：%s' % (time.time() - start)