VNPY批量优化参数，并输出到excel

# encoding: UTF-8

import pandas as pd

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

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

class BatchOptimization(object):

    def __init__(self):

        ""

    def calculateBacktesting(self,symbollist,strategylist, sort = 'totalNetPnl'):

        #填入品种队列和策略队列，返回结果resultlist, 为了输出方便检索，加入品种名称，策略名称和策略参数

        resultlist = []

        for symbol in symbollist:

            for strategy in strategylist:

                result = self.runBacktesting(symbol,strategy,sort)

                #加入品种名称，策略名称和策略参数

                if isinstance(result,dict):

                    #如果返回的是dict，直接加入

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

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

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

                    resultlist.append(result)

                else:

                    # 发现优化回来的是一个包含元组的队列，元组有三个组成，第一个排策略参数，第二个回测目标的值，第三策略参数全部运行结果。

                    # 这里我们要的就是第三个,先插入这个dict，在dict插入symbolname，和strategysetting

                    for resultraw in result:

                        resultlist.append(resultraw[2])

                        resultlist[-1]["Symbolname"] = str(symbol["vtSymbol"])

                        resultlist[-1]["strategysetting"] = str(resultraw[0])

        return resultlist

    def runBacktesting(self, symbol, strategy, sort):

        #写入测试品种和参数, 返回回测数据集包含回测结果

        # 在引擎中创建策略对象

        # 创建回测引擎

        engine = BacktestingEngine()

        # 设置引擎的回测模式为K线

        engine.setBacktestingMode(engine.BAR_MODE)

        # 设置回测用的数据起始日期

        engine.setStartDate(symbol["StartDate"])

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

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

        engine.setSize(symbol["Size"]) # 合约大小

        engine.setPriceTick(symbol["Slippage"]) # 最小价格变动

        engine.setCapital(symbol["Capital"])

        # 设置使用的历史数据库

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

        #调用优化方法，可以集成优化测试

        setting = OptimizationSetting() # 新建一个优化任务设置对象

        setting.setOptimizeTarget(sort) # 设置优化排序的目标是策略净盈利

        print strategy[1]

        for settingKey in strategy[1]:

            if isinstance(strategy[1][settingKey], tuple):

                setting.addParameter(settingKey,strategy[1][settingKey][0],strategy[1][settingKey][1],strategy[1][settingKey][2])

            else:

                setting.addParameter(settingKey,strategy[1][settingKey])

        #

        optimizationresult = engine.runParallelOptimization(strategy[0], setting)

        engine.output(u'输出统计数据')

        # 如果是使用优化模式，这里返回的是策略回测的dict的list，如果普通回测就是单个dict

        # 如果大于30 ，就返回三十之内，否则全部

        if len(optimizationresult) > 30:

            return optimizationresult[:30]

        else:

            return optimizationresult

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

        #按照输入统计数据队列和路径，输出excel，这里不提供新增模式，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 "回测统计结果输出到" + path

if __name__ == "__main__":

    #创建品种队列，这里可以用json导入，为了方便使用直接写了。

    symbollist = [{

        "vtSymbol": 'm1809',

        "StartDate": "20180101",

        "Slippage": 1,

        "Size": 10,

        "Rate": 2 / 10000,

        "Capital": 10000

            },

        {

            "vtSymbol": 'rb1810',

            "StartDate": "20180101",

            "Slippage": 1,

            "Size": 10,

            "Rate": 2 / 10000,

            "Capital": 10000

        }

    ]

    # 这里是同一个策略，不同参数的情况，当然可以有多个策略和多个参数组合

    Strategylist2 = []

    # 策略list,如果是元组，那么就是三个，按照第一个初始，第二个结束，第三个步进

    settinglist =[

        {'bollWindow': (10,20,2)}]

    # 合并一个元组

    if settinglist != []:

        for para1 in settinglist:

            Strategylist2.append((BollChannelStrategy, para1))

    NT = BatchOptimization()

    resultlist = NT.calculateBacktesting(symbollist,Strategylist2,sort = 'totalNetPnl')

    #定义路径

    path = "C:\Project\OptimizationResult.xlsx"

    NT.toExcel(resultlist,path)