请 [注册] 或 [登录]  | 返回主站

量化交易吧 /  量化策略 帖子:3351420 新帖:38

扒一扒常见技术指标的大盘择时效果

舞蹈家2发表于:5 月 10 日 16:59回复(2)

技术指标说明

  • 说到技术指标,想必不论新韭老韭都能随口念出几个来,是判断行情走势的衍生工具,也是行情软件中必备的板块内容,技术分析一直在大多数股民中备受好评,各个频道的股评专家也是乐此不疲。
    股市参与者千万人,在同一市场中博弈,作为个人投资者获取行业或公司信息并无优势,然而关于市场所有的信息最终都会落到交易上,会体现在成交量和价格上,量价信息这个层面来讲,市场是公平的,这也是技术分析能够为大多数人使用的原因之一。
    zhib.png

  • 使用技术分析之前,我们需要了解到,所有技术分析能够成立,是建立在以下的三大基石之上

    1. 市场行为包容消化一切信息
    2. 价格以趋势方式演变
    3. 历史会重演
  • 基于以上假设,演化出了各类市场分析的手段方法,下面进行了简单的汇总,技术指标作为技术分析的一个分支,其方便、直观、快捷的使用方式很受投资者的欢迎,其中蕴涵了投资者对股价变化进行长期观察并积累经验,逐步归纳总结出来的有关股市波动的规律,本篇旨在针对技术指标进行一些探索。技术分析.png

进行大盘择时效果检查

  • 技术指标关注的点在于标的在市场中的成交量与价格变化情况,并没有去考虑所参与的标的是造酒的还是卖软件的,我们有理由认为股价变动会与自身的实际业务及行业而具有自己独特的规律,所以,这里为了避免在个股操作中受到行业及个股自身特性的干扰,我们直接拿指标进行指数交易。这里我们选了常见的13个技术指标,指标列表下
    [(0, 'None'),(1, 'MA'),(2, 'MA1'),(3, 'EMA'),(4, 'EMA1'),(5, 'MACD'),(6, 'KDJ'),(7, 'CCI'),(8, 'RSI'),(9, 'BOLL'),(10, 'BIAS'),(11, 'TRIX'),(12, 'BBI'),(13, 'PSY')]
    
    注:后面的分指标回测图,数字与这里的指标是对应的关系
  • 用以上指标分别统计了各个指标在沪深300、上证50、中证500
    、中证1000指数上的应用效果,为了能对结果有较为直观的对比,这里加入了指标‘None’,用于记录全仓买入指数没有择时的情况
  • 下面是指标在沪深300指数上的择时表现,以按收益进行排序

    表格图1
    d27ac93238c2d1b99ec6d968e66ac782

    可以看到13个指标,有9个是可以跑赢大盘本身,排名第一的是反转类指标(一般均线突破类的为趋势类,设置临界阈值类的指标为反转类),且大部分回撤都集中在大盘股灾期间

    注:这里旨在进行指标效果探索,已将手续费和滑点设置为0
  • 下面是指标在上证50指数上的择时表现
    表格图1 d27ac93238c2d1b99ec6d968e66ac782

    其中只有3个指标能够跑赢大盘,排名第一的是趋势类指标布林带,大部分回撤集中在08年股灾后到15年股灾前的长熊期间
  • 下面是指标在中证500指数上的择时表现
    表格图1 d27ac93238c2d1b99ec6d968e66ac782

    其中有4个指标能跑赢大盘,排名第一的是趋势类指标指数移动平均线,大部分回撤都集中在15年股灾后
  • 下图是指数在中证1000指数上的择时表现
    表格图1 d27ac93238c2d1b99ec6d968e66ac782

    中证1000由于指数编制较晚,出现了较长时间的空仓期,10个指标是可以跑赢指数本身的,排名第一的是趋势类布林带指标,BOLL指标第二次出现,回撤集中在15年股灾后。

参数调试

  • 以上指标进行测试直接调用的聚宽技术指标库的方法,涉及到指标参数均使用已设置的默认参数,部分涉及到设置信号阈值类的指标,如RSI指标,目前是主观给的范围(未做调优),为了进一步测试指标的效果是否稳定(Robust),下面将进行策略收益对参数变化的敏感性测试,我们取了两个指标进行测试,分别是在中证500上效果显著的EMA指数移动平均线指标,和在中证1000上效果显著的BOLL带指标,设置不同统计长度(timeperiod),分别设置为
    ['10','11','12','13','14','15','16','17','18','19','20','21','22','23','24','25','26','27','28','30']
    
    敏感度.pngboll.png
  • 可以看到,EMA表现最好的参数为21,BOLL带表现最好的指标是23,并且在两边是逐步下降的过程,而实际上,指标表现最好的参数基本就是一个月的交易日个数,我们猜测市场存在明显的月度效应(这里可以试着拿该指标在更多的标的中进行测试,分别统计不同参数下,跑赢不择时的比率变化规律)。

结果说明

通过上述的测试结果,我们初步发现如下的情况

  • 在4个指数上都能跑赢指数本身的指标有两个,MA、RSI(这里不禁感慨难怪信仰均线的人不在少数,确实是广泛有效的指标)
  • 均线类指标统计长度(timeperiod)变现最好的参数基本就是一个月交易日个数
  • 择时指标开平仓普遍胜率不高,由此推测盈利主要来自于盈亏比
  • 趋势类指标效果整体优于反转类指标

本篇为了避免个别标的自身特性的影响,都是拿指数进行的测试,然而不同指数本身也有各自的特点,可以借助已有的代码进行更多标的枚举暴力测试,通过多回测方法进一步挖掘更为有效的指标参数。

20181225:感谢@宋一堆 的改进思路,已将可用做空版代码附在评论区
20181225:感谢评论区sailer 和 iamrobot 的留言,文章统计结果已替换为修正后的结果,代码已更改附在评论区


#1 先导入所需要的程序包
import datetime
import numpy as np
import pandas as pd
import time
from jqdata import *
from pandas import Series, DataFrame
import matplotlib.pyplot as plt
import seaborn as sns
import itertools
import copy
import pickle

# 定义类'参数分析'
class parameter_analysis(object):
    
    # 定义函数中不同的变量
    def __init__(self, algorithm_id=None):
        self.algorithm_id = algorithm_id            # 回测id
        
        self.params_df = pd.DataFrame()             # 回测中所有调参备选值的内容,列名字为对应修改面两名称,对应回测中的 g.XXXX
        self.results = {}                           # 回测结果的回报率,key 为 params_df 的行序号,value 为
        self.evaluations = {}                       # 回测结果的各项指标,key 为 params_df 的行序号,value 为一个 dataframe
        self.backtest_ids = {}                      # 回测结果的 id
        
        # 新加入的基准的回测结果 id,可以默认为空 '',则使用回测中设定的基准
        self.benchmark_id = None                      
        
        self.benchmark_returns = []                 # 新加入的基准的回测回报率
        self.returns = {}                           # 记录所有回报率
        self.excess_returns = {}                    # 记录超额收益率
        self.log_returns = {}                       # 记录收益率的 log 值
        self.log_excess_returns = {}                # 记录超额收益的 log 值
        self.dates = []                             # 回测对应的所有日期
        self.excess_max_drawdown = {}               # 计算超额收益的最大回撤
        self.excess_annual_return = {}              # 计算超额收益率的年化指标
        self.evaluations_df = pd.DataFrame()        # 记录各项回测指标,除日回报率外
    
    # 定义排队运行多参数回测函数
    def run_backtest(self,                          #
                     algorithm_id=None,             # 回测策略id
                     running_max=10,                # 回测中同时巡行最大回测数量
                     start_date='2006-01-01',       # 回测的起始日期
                     end_date='2016-11-30',         # 回测的结束日期
                     frequency='day',               # 回测的运行频率
                     initial_cash='1000000',        # 回测的初始持仓金额
                     param_names=[],                # 回测中调整参数涉及的变量
                     param_values=[]                # 回测中每个变量的备选参数值
                     ):
        # 当此处回测策略的 id 没有给出时,调用类输入的策略 id
        if algorithm_id == None: algorithm_id=self.algorithm_id
        
        # 生成所有参数组合并加载到 df 中
        # 包含了不同参数具体备选值的排列组合中一组参数的 tuple 的 list
        param_combinations = list(itertools.product(*param_values))
        # 生成一个 dataframe, 对应的列为每个调参的变量,每个值为调参对应的备选值
        to_run_df = pd.DataFrame(param_combinations)
        # 修改列名称为调参变量的名字
        to_run_df.columns = param_names
        
        # 设定运行起始时间和保存格式
        start = time.time()
        # 记录结束的运行回测
        finished_backtests = {}
        # 记录运行中的回测
        running_backtests = {}
        # 计数器
        pointer = 0
        # 总运行回测数目,等于排列组合中的元素个数
        total_backtest_num = len(param_combinations)
        # 记录回测结果的回报率
        all_results = {}
        # 记录回测结果的各项指标
        all_evaluations = {}
        
        # 在运行开始时显示
        print ('【已完成|运行中|待运行】:'), 
        # 当运行回测开始后,如果没有全部运行完全的话:
        while len(finished_backtests)<total_backtest_num:
            # 显示运行、完成和待运行的回测个数
            print('[%s|%s|%s].' % (len(finished_backtests), 
                                   len(running_backtests), 
                                   (total_backtest_num-len(finished_backtests)-len(running_backtests)) )),
            # 记录当前运行中的空位数量
            to_run = min(running_max-len(running_backtests), total_backtest_num-len(running_backtests)-len(finished_backtests))
            # 把可用的空位进行跑回测
            for i in range(pointer, pointer+to_run):
                # 备选的参数排列组合的 df 中第 i 行变成 dict,每个 key 为列名字,value 为 df 中对应的值
                params = to_run_df.iloc[i].to_dict()
                # 记录策略回测结果的 id,调整参数 extras 使用 params 的内容
                backtest = create_backtest(algorithm_id = algorithm_id,
                                           start_date = start_date, 
                                           end_date = end_date, 
                                           frequency = frequency, 
                                           initial_cash = initial_cash, 
                                           extras = params, 
                                           # 再回测中把改参数的结果起一个名字,包含了所有涉及的变量参数值
                                           name = str(params)
                                           )
                # 记录运行中 i 回测的回测 id
                running_backtests[i] = backtest
            # 计数器计数运行完的数量    
            pointer = pointer+to_run
            
            # 获取回测结果
            failed = []
            finished = []
            # 对于运行中的回测,key 为 to_run_df 中所有排列组合中的序数
            for key in running_backtests.keys():
                # 研究调用回测的结果,running_backtests[key] 为运行中保存的结果 id
                bt = get_backtest(running_backtests[key])
                # 获得运行回测结果的状态,成功和失败都需要运行结束后返回,如果没有返回则运行没有结束
                status = bt.get_status()
                # 当运行回测失败
                if status == 'failed':
                    # 失败 list 中记录对应的回测结果 id
                    failed.append(key)
                # 当运行回测成功时
                elif status == 'done':
                    # 成功 list 记录对应的回测结果 id,finish 仅记录运行成功的
                    finished.append(key)
                    # 回测回报率记录对应回测的回报率 dict, key to_run_df 中所有排列组合中的序数, value 为回报率的 dict
                    # 每个 value 一个 list 每个对象为一个包含时间、日回报率和基准回报率的 dict
                    all_results[key] = bt.get_results()
                    # 回测回报率记录对应回测结果指标 dict, key to_run_df 中所有排列组合中的序数, value 为回测结果指标的 dataframe
                    all_evaluations[key] = bt.get_risk()
            # 记录运行中回测结果 id 的 list 中删除失败的运行
            for key in failed:
                running_backtests.pop(key)
            # 在结束回测结果 dict 中记录运行成功的回测结果 id,同时在运行中的记录中删除该回测
            for key in finished:
                finished_backtests[key] = running_backtests.pop(key)
            # 当一组同时运行的回测结束时报告时间
            if len(finished_backtests) != 0 and len(finished_backtests) % running_max == 0 and to_run !=0:
                # 记录当时时间
                middle = time.time()
                # 计算剩余时间,假设没工作量时间相等的话
                remain_time = (middle - start) * (total_backtest_num - len(finished_backtests)) / len(finished_backtests)
                # print 当前运行时间
                print('[已用%s时,尚余%s时,请不要关闭浏览器].' % (str(round((middle - start) / 60.0 / 60.0,3)), 
                                          str(round(remain_time / 60.0 / 60.0,3)))),
            # 5秒钟后再跑一下
            time.sleep(30) 
        # 记录结束时间
        end = time.time() 
        print ('')
        print('【回测完成】总用时:%s秒(即%s小时)。' % (str(int(end-start)), 
                                           str(round((end-start)/60.0/60.0,2)))),
        # 对应修改类内部对应
        self.params_df = to_run_df
        self.results = all_results
        self.evaluations = all_evaluations
        self.backtest_ids = finished_backtests

        
    #7 最大回撤计算方法
    def find_max_drawdown(self, returns):
        # 定义最大回撤的变量
        result = 0
        # 记录最高的回报率点
        historical_return = 0
        # 遍历所有日期
        for i in range(len(returns)):
            # 最高回报率记录
            historical_return = max(historical_return, returns[i])
            # 最大回撤记录
            drawdown = 1-(returns[i] + 1) / (historical_return + 1)
            # 记录最大回撤
            result = max(drawdown, result)
        # 返回最大回撤值
        return result

    # log 收益、新基准下超额收益和相对与新基准的最大回撤
    def organize_backtest_results(self, benchmark_id=None):
        # 若新基准的回测结果 id 没给出
        if benchmark_id==None:
            # 使用默认的基准回报率,默认的基准在回测策略中设定
            self.benchmark_returns = [x['benchmark_returns'] for x in self.results[0]]
        # 当新基准指标给出后    
        else:
            # 基准使用新加入的基准回测结果
            self.benchmark_returns = [x['returns'] for x in get_backtest(benchmark_id).get_results()]
        # 回测日期为结果中记录的第一项对应的日期
        self.dates = [x['time'] for x in self.results[0]]
        
        # 对应每个回测在所有备选回测中的顺序 (key),生成新数据
        # 由 {key:{u'benchmark_returns': 0.022480100091729405,
        #           u'returns': 0.03184566700000002,
        #           u'time': u'2006-02-14'}} 格式转化为:
        # {key: []} 格式,其中 list 为对应 date 的一个回报率 list
        for key in self.results.keys():
            self.returns[key] = [x['returns'] for x in self.results[key]]
        # 生成对于基准(或新基准)的超额收益率
        for key in self.results.keys():
            self.excess_returns[key] = [(x+1)/(y+1)-1 for (x,y) in zip(self.returns[key], self.benchmark_returns)]
        # 生成 log 形式的收益率
        for key in self.results.keys():
            self.log_returns[key] = [log(x+1) for x in self.returns[key]]
        # 生成超额收益率的 log 形式
        for key in self.results.keys():
            self.log_excess_returns[key] = [log(x+1) for x in self.excess_returns[key]]
        # 生成超额收益率的最大回撤
        for key in self.results.keys():
            self.excess_max_drawdown[key] = self.find_max_drawdown(self.excess_returns[key])
        # 生成年化超额收益率
        for key in self.results.keys():
            self.excess_annual_return[key] = (self.excess_returns[key][-1]+1)**(252./float(len(self.dates)))-1
        # 把调参数据中的参数组合 df 与对应结果的 df 进行合并
        self.evaluations_df = pd.concat([self.params_df, pd.DataFrame(self.evaluations).T], axis=1)
#         self.evaluations_df = 

    # 获取最总分析数据,调用排队回测函数和数据整理的函数    
    def get_backtest_data(self,
                          algorithm_id=None,                         # 回测策略id
                          benchmark_id=None,                         # 新基准回测结果id
                          file_name='results.pkl',                   # 保存结果的 pickle 文件名字
                          running_max=10,                            # 最大同时运行回测数量
                          start_date='2006-01-01',                   # 回测开始时间
                          end_date='2016-11-30',                     # 回测结束日期
                          frequency='day',                           # 回测的运行频率
                          initial_cash='1000000',                    # 回测初始持仓资金
                          param_names=[],                            # 回测需要测试的变量
                          param_values=[]                            # 对应每个变量的备选参数
                          ):
        # 调运排队回测函数,传递对应参数
        self.run_backtest(algorithm_id=algorithm_id,
                          running_max=running_max,
                          start_date=start_date,
                          end_date=end_date,
                          frequency=frequency,
                          initial_cash=initial_cash,
                          param_names=param_names,
                          param_values=param_values
                          )
        # 回测结果指标中加入 log 收益率和超额收益率等指标
        self.organize_backtest_results(benchmark_id)
        # 生成 dict 保存所有结果。
        results = {'returns':self.returns,
                   'excess_returns':self.excess_returns,
                   'log_returns':self.log_returns,
                   'log_excess_returns':self.log_excess_returns,
                   'dates':self.dates,
                   'benchmark_returns':self.benchmark_returns,
                   'evaluations':self.evaluations,
                   'params_df':self.params_df,
                   'backtest_ids':self.backtest_ids,
                   'excess_max_drawdown':self.excess_max_drawdown,
                   'excess_annual_return':self.excess_annual_return,
                   'evaluations_df':self.evaluations_df}
        # 保存 pickle 文件
        pickle_file = open(file_name, 'wb')
        pickle.dump(results, pickle_file)
        pickle_file.close()

    # 读取保存的 pickle 文件,赋予类中的对象名对应的保存内容    
    def read_backtest_data(self, file_name='results.pkl'):
        pickle_file = open(file_name, 'rb')
        results = pickle.load(pickle_file)
        self.returns = results['returns']
        self.excess_returns = results['excess_returns']
        self.log_returns = results['log_returns']
        self.log_excess_returns = results['log_excess_returns']
        self.dates = results['dates']
        self.benchmark_returns = results['benchmark_returns']
        self.evaluations = results['evaluations']
        self.params_df = results['params_df']
        self.backtest_ids = results['backtest_ids']
        self.excess_max_drawdown = results['excess_max_drawdown']
        self.excess_annual_return = results['excess_annual_return']
        self.evaluations_df = results['evaluations_df']
        
    # 回报率折线图    
    def plot_returns(self):
        # 通过figsize参数可以指定绘图对象的宽度和高度,单位为英寸;
        fig = plt.figure(figsize=(20,8))
        ax = fig.add_subplot(111)
        # 作图
        for key in self.returns.keys():
            ax.plot(range(len(self.returns[key])), self.returns[key], label=key)
        # 设定benchmark曲线并标记
        ax.plot(range(len(self.benchmark_returns)), self.benchmark_returns, label='benchmark', c='k', linestyle='--') 
        ticks = [int(x) for x in np.linspace(0, len(self.dates)-1, 11)]
        plt.xticks(ticks, [self.dates[i] for i in ticks])
        # 设置图例样式
        ax.legend(loc = 2, fontsize = 10)
        # 设置y标签样式
        ax.set_ylabel('returns',fontsize=20)
        # 设置x标签样式
        ax.set_yticklabels([str(x*100)+'% 'for x in ax.get_yticks()])
        # 设置图片标题样式
        ax.set_title("Strategy's performances with different parameters", fontsize=21)
        plt.xlim(0, len(self.returns[0]))

    # 超额收益率图    
    def plot_excess_returns(self):
        # 通过figsize参数可以指定绘图对象的宽度和高度,单位为英寸;
        fig = plt.figure(figsize=(20,8))
        ax = fig.add_subplot(111)
        # 作图
        for key in self.returns.keys():
            ax.plot(range(len(self.excess_returns[key])), self.excess_returns[key], label=key)
        # 设定benchmark曲线并标记
        ax.plot(range(len(self.benchmark_returns)), [0]*len(self.benchmark_returns), label='benchmark', c='k', linestyle='--')
        ticks = [int(x) for x in np.linspace(0, len(self.dates)-1, 11)]
        plt.xticks(ticks, [self.dates[i] for i in ticks])
        # 设置图例样式
        ax.legend(loc = 2, fontsize = 10)
        # 设置y标签样式
        ax.set_ylabel('excess returns',fontsize=20)
        # 设置x标签样式
        ax.set_yticklabels([str(x*100)+'% 'for x in ax.get_yticks()])
        # 设置图片标题样式
        ax.set_title("Strategy's performances with different parameters", fontsize=21)
        plt.xlim(0, len(self.excess_returns[0]))
        
    # log回报率图    
    def plot_log_returns(self):
        # 通过figsize参数可以指定绘图对象的宽度和高度,单位为英寸;
        fig = plt.figure(figsize=(20,8))
        ax = fig.add_subplot(111)
        # 作图
        for key in self.returns.keys():
            ax.plot(range(len(self.log_returns[key])), self.log_returns[key], label=key)
        # 设定benchmark曲线并标记
        ax.plot(range(len(self.benchmark_returns)), [log(x+1) for x in self.benchmark_returns], label='benchmark', c='k', linestyle='--')
        ticks = [int(x) for x in np.linspace(0, len(self.dates)-1, 11)]
        plt.xticks(ticks, [self.dates[i] for i in ticks])
        # 设置图例样式
        ax.legend(loc = 2, fontsize = 10)
        # 设置y标签样式
        ax.set_ylabel('log returns',fontsize=20)
        # 设置图片标题样式
        ax.set_title("Strategy's performances with different parameters", fontsize=21)
        plt.xlim(0, len(self.log_returns[0]))
    
    # 超额收益率的 log 图
    def plot_log_excess_returns(self):
        # 通过figsize参数可以指定绘图对象的宽度和高度,单位为英寸;
        fig = plt.figure(figsize=(20,8))
        ax = fig.add_subplot(111)
        # 作图
        for key in self.returns.keys():
            ax.plot(range(len(self.log_excess_returns[key])), self.log_excess_returns[key], label=key)
        # 设定benchmark曲线并标记
        ax.plot(range(len(self.benchmark_returns)), [0]*len(self.benchmark_returns), label='benchmark', c='k', linestyle='--')
        ticks = [int(x) for x in np.linspace(0, len(self.dates)-1, 11)]
        plt.xticks(ticks, [self.dates[i] for i in ticks])
        # 设置图例样式
        ax.legend(loc = 2, fontsize = 10)
        # 设置y标签样式
        ax.set_ylabel('log excess returns',fontsize=20)
        # 设置图片标题样式
        ax.set_title("Strategy's performances with different parameters", fontsize=21)
        plt.xlim(0, len(self.log_excess_returns[0]))

        
    # 回测的4个主要指标,包括总回报率、最大回撤夏普率和波动
    def get_eval4_bar(self, sort_by=[]): 
        
        sorted_params = self.params_df
        for by in sort_by:
            sorted_params = sorted_params.sort(by)
        indices = sorted_params.index
        
        fig = plt.figure(figsize=(20,7))

        # 定义位置
        ax1 = fig.add_subplot(221)
        # 设定横轴为对应分位,纵轴为对应指标
        ax1.bar(range(len(indices)), 
                [self.evaluations[x]['algorithm_return'] for x in indices], 0.6, label = 'Algorithm_return')
        plt.xticks([x+0.3 for x in range(len(indices))], indices)
        # 设置图例样式
        ax1.legend(loc='best',fontsize=15)
        # 设置y标签样式
        ax1.set_ylabel('Algorithm_return', fontsize=15)
        # 设置y标签样式
        ax1.set_yticklabels([str(x*100)+'% 'for x in ax1.get_yticks()])
        # 设置图片标题样式
        ax1.set_title("Strategy's of Algorithm_return performances of different quantile", fontsize=15)
        # x轴范围
        plt.xlim(0, len(indices))

        # 定义位置
        ax2 = fig.add_subplot(224)
        # 设定横轴为对应分位,纵轴为对应指标
        ax2.bar(range(len(indices)), 
                [self.evaluations[x]['max_drawdown'] for x in indices], 0.6, label = 'Max_drawdown')
        plt.xticks([x+0.3 for x in range(len(indices))], indices)
        # 设置图例样式
        ax2.legend(loc='best',fontsize=15)
        # 设置y标签样式
        ax2.set_ylabel('Max_drawdown', fontsize=15)
        # 设置x标签样式
        ax2.set_yticklabels([str(x*100)+'% 'for x in ax2.get_yticks()])
        # 设置图片标题样式
        ax2.set_title("Strategy's of Max_drawdown performances of different quantile", fontsize=15)
        # x轴范围
        plt.xlim(0, len(indices))

        # 定义位置
        ax3 = fig.add_subplot(223)
        # 设定横轴为对应分位,纵轴为对应指标
        ax3.bar(range(len(indices)),
                [self.evaluations[x]['sharpe'] for x in indices], 0.6, label = 'Sharpe')
        plt.xticks([x+0.3 for x in range(len(indices))], indices)
        # 设置图例样式
        ax3.legend(loc='best',fontsize=15)
        # 设置y标签样式
        ax3.set_ylabel('Sharpe', fontsize=15)
        # 设置x标签样式
        ax3.set_yticklabels([str(x*100)+'% 'for x in ax3.get_yticks()])
        # 设置图片标题样式
        ax3.set_title("Strategy's of Sharpe performances of different quantile", fontsize=15)
        # x轴范围
        plt.xlim(0, len(indices))

        # 定义位置
        ax4 = fig.add_subplot(222)
        # 设定横轴为对应分位,纵轴为对应指标
        ax4.bar(range(len(indices)), 
                [self.evaluations[x]['algorithm_volatility'] for x in indices], 0.6, label = 'Algorithm_volatility')
        plt.xticks([x+0.3 for x in range(len(indices))], indices)
        # 设置图例样式
        ax4.legend(loc='best',fontsize=15)
        # 设置y标签样式
        ax4.set_ylabel('Algorithm_volatility', fontsize=15)
        # 设置x标签样式
        ax4.set_yticklabels([str(x*100)+'% 'for x in ax4.get_yticks()])
        # 设置图片标题样式
        ax4.set_title("Strategy's of Algorithm_volatility performances of different quantile", fontsize=15)
        # x轴范围
        plt.xlim(0, len(indices))
        
    #14 年化回报和最大回撤,正负双色表示
    def get_eval(self, sort_by=[]):

        sorted_params = self.params_df
        for by in sort_by:
            sorted_params = sorted_params.sort(by)
        indices = sorted_params.index
        
        # 大小
        fig = plt.figure(figsize = (20, 8))
        # 图1位置
        ax = fig.add_subplot(111)
        # 生成图超额收益率的最大回撤
        ax.bar([x+0.3 for x in range(len(indices))],
               [-self.evaluations[x]['max_drawdown'] for x in indices], color = '#32CD32',  
                     width = 0.6, label = 'Max_drawdown', zorder=10)
        # 图年化超额收益
        ax.bar([x for x in range(len(indices))],
               [self.evaluations[x]['annual_algo_return'] for x in indices], color = 'r', 
                     width = 0.6, label = 'Annual_return')
        plt.xticks([x+0.3 for x in range(len(indices))], indices)
        # 设置图例样式
        ax.legend(loc='best',fontsize=15)
        # 基准线
        plt.plot([0, len(indices)], [0, 0], c='k', 
                 linestyle='--', label='zero')
        # 设置图例样式
        ax.legend(loc='best',fontsize=15)
        # 设置y标签样式
        ax.set_ylabel('Max_drawdown', fontsize=15)
        # 设置x标签样式
        ax.set_yticklabels([str(x*100)+'% 'for x in ax.get_yticks()])
        # 设置图片标题样式
        ax.set_title("Strategy's performances of different quantile", fontsize=15)
        #   设定x轴长度
        plt.xlim(0, len(indices))


    #14 超额收益的年化回报和最大回撤
    # 加入新的benchmark后超额收益和
    def get_excess_eval(self, sort_by=[]):

        sorted_params = self.params_df
        for by in sort_by:
            sorted_params = sorted_params.sort(by)
        indices = sorted_params.index
        
        # 大小
        fig = plt.figure(figsize = (20, 8))
        # 图1位置
        ax = fig.add_subplot(111)
        # 生成图超额收益率的最大回撤
        ax.bar([x+0.3 for x in range(len(indices))],
               [-self.excess_max_drawdown[x] for x in indices], color = '#32CD32',  
                     width = 0.6, label = 'Excess_max_drawdown')
        # 图年化超额收益
        ax.bar([x for x in range(len(indices))],
               [self.excess_annual_return[x] for x in indices], color = 'r', 
                     width = 0.6, label = 'Excess_annual_return')
        plt.xticks([x+0.3 for x in range(len(indices))], indices)
        # 设置图例样式
        ax.legend(loc='best',fontsize=15)
        # 基准线
        plt.plot([0, len(indices)], [0, 0], c='k', 
                 linestyle='--', label='zero')
        # 设置图例样式
        ax.legend(loc='best',fontsize=15)
        # 设置y标签样式
        ax.set_ylabel('Max_drawdown', fontsize=15)
        # 设置x标签样式
        ax.set_yticklabels([str(x*100)+'% 'for x in ax.get_yticks()])
        # 设置图片标题样式
        ax.set_title("Strategy's performances of different quantile", fontsize=15)
        #   设定x轴长度
        plt.xlim(0, len(indices))

应用多策略回测框架,统计回测结果¶

  • 根据回测algorithmId在研究中创建策略
  • 设置不同指标的回测组别
  • 读取回测结果进行展示
#2 设定回测策略 id 
# 注意!注意!注意!这里的id是在 我的策略里面的编译运行的algorithmId,在浏览器地址里面复制一下
pa = parameter_analysis('5473a8b9e878d36100307a7d1c0405ee')
#3 运行回测
pa.get_backtest_data(file_name = 'results1.pkl',
                          running_max = 10,
                          benchmark_id = None,
                          start_date = '2008-12-05',
                           end_date = '2018-12-05',
                          frequency = 'day',
                          initial_cash = '100000000',
                          param_names = ['tech'],
                          param_values = [['None','MA','MA1','EMA','EMA1','MACD','KDJ','CCI','RSI','BOLL','BIAS','TRIX','BBI','PSY']]
                          )
【已完成|运行中|待运行】: [0|0|14]. [0|10|4]. [0|10|4]. [8|2|4]. [已用0.031时,尚余0.012时,请不要关闭浏览器]. [10|4|0]. [10|4|0]. 
【回测完成】总用时:206秒(即0.06小时)。
#4 数据读取
pa.read_backtest_data('results1.pkl')
#5 查看回测参数的df
pa.params_df
.dataframe thead tr:only-child th { text-align: right; } .dataframe thead th { text-align: left; } .dataframe tbody tr th { vertical-align: top; }
tech
0 None
1 MA
2 MA1
3 EMA
4 EMA1
5 MACD
6 KDJ
7 CCI
8 RSI
9 BOLL
10 BIAS
11 TRIX
12 BBI
13 PSY
#6 查看回测结果指标
df = pa.evaluations_df
df
.dataframe thead tr:only-child th { text-align: right; } .dataframe thead th { text-align: left; } .dataframe tbody tr th { vertical-align: top; }
tech __version algorithm_return algorithm_volatility alpha annual_algo_return annual_bm_return avg_position_days avg_trade_return benchmark_return ... max_drawdown_period max_leverage period_label profit_loss_ratio sharpe sortino trading_days treasury_return win_count win_ratio
0 None 101 0.649806 0.246126 0.000651047 0.0527908 0.052146 2433 0 0.639999 ... [2015-06-08, 2016-01-28] 0 2018-12 0 0.0519684 0.0660195 2433 0.400219 0 0
1 MA 101 1.16138 0.158064 0.0374493 0.0824177 0.052146 1297 0.00877665 0.639999 ... [2015-06-08, 2018-12-05] 0 2018-12 1.52665 0.268359 0.351814 2433 0.400219 37 0.308333
2 MA1 101 0.362528 0.168506 -0.0133998 0.0322966 0.052146 1307 0.00797548 0.639999 ... [2015-06-08, 2018-11-26] 0 2018-12 1.17302 -0.0457156 -0.0566109 2433 0.400219 30 0.405405
3 EMA 101 1.07912 0.157455 0.0331647 0.0781108 0.052146 1305 0.00859248 0.639999 ... [2009-08-03, 2010-03-25] 0 2018-12 1.62981 0.242043 0.317869 2433 0.400219 34 0.285714
4 EMA1 101 0.848388 0.167936 0.0194674 0.0651579 0.052146 1285 0.0272402 0.639999 ... [2009-08-03, 2014-07-10] 0 2018-12 1.59803 0.149806 0.189869 2433 0.400219 10 0.25641
5 MACD 101 -0.248712 0.186899 -0.075951 -0.0289566 0.052146 1157 -0.000632834 0.639999 ... [2009-08-03, 2018-08-06] 0 2018-12 0.90209 -0.368951 -0.404536 2433 0.400219 72 0.679245
6 KDJ 101 1.19036 0.187656 0.0368768 0.0839002 0.052146 1176 0.00491715 0.639999 ... [2015-06-03, 2015-08-26] 0 2018-12 1.3062 0.233939 0.297756 2433 0.400219 161 0.715556
7 CCI 101 0.297372 0.165 -0.0183251 0.027112 0.052146 1296 0.00630609 0.639999 ... [2015-06-08, 2018-12-05] 0 2018-12 1.17069 -0.078109 -0.09669 2433 0.400219 32 0.405063
8 RSI 101 1.03938 0.163683 0.0305914 0.0759747 0.052146 1247 0.0447027 0.639999 ... [2010-11-08, 2014-04-29] 0 2018-12 1.87224 0.219782 0.284327 2433 0.400219 9 0.346154
9 BOLL 101 1.10888 0.125783 0.0364571 0.0796862 0.052146 790 0.0216967 0.639999 ... [2015-06-08, 2016-11-02] 0 2018-12 2.20369 0.315514 0.429048 2433 0.400219 19 0.452381
10 BIAS 101 1.16048 0.169097 0.0366637 0.0823712 0.052146 1258 0.016247 0.639999 ... [2015-06-08, 2015-09-15] 0 2018-12 1.43307 0.250573 0.319445 2433 0.400219 30 0.454545
11 TRIX 101 1.16048 0.169097 0.0366637 0.0823712 0.052146 1258 0.016247 0.639999 ... [2015-06-08, 2015-09-15] 0 2018-12 1.43307 0.250573 0.319445 2433 0.400219 30 0.454545
12 BBI 101 0.858419 0.155386 0.0208905 0.0657505 0.052146 1278 0.00460358 0.639999 ... [2015-05-26, 2018-12-05] 0 2018-12 1.32596 0.165719 0.221382 2433 0.400219 69 0.334951
13 PSY 101 0.335471 0.164411 -0.015234 0.0301712 0.052146 1045 0.011718 0.639999 ... [2010-03-03, 2012-01-05] 0 2018-12 1.4024 -0.0597818 -0.0815307 2433 0.400219 18 0.545455

14 rows × 27 columns

#7 回报率折线图    
pa.plot_returns()
#8 超额收益率图    
pa.plot_excess_returns()
#指标回测收益列表
df.index = df['tech'].values
del df['tech']
df = df[['algorithm_return','alpha','sharpe','win_ratio','max_drawdown_period']]
df.sort_values('algorithm_return',ascending=0)
.dataframe thead tr:only-child th { text-align: right; } .dataframe thead th { text-align: left; } .dataframe tbody tr th { vertical-align: top; }
algorithm_return alpha sharpe win_ratio max_drawdown_period
KDJ 1.19036 0.0368768 0.233939 0.715556 [2015-06-03, 2015-08-26]
MA 1.16138 0.0374493 0.268359 0.308333 [2015-06-08, 2018-12-05]
BIAS 1.16048 0.0366637 0.250573 0.454545 [2015-06-08, 2015-09-15]
TRIX 1.16048 0.0366637 0.250573 0.454545 [2015-06-08, 2015-09-15]
BOLL 1.10888 0.0364571 0.315514 0.452381 [2015-06-08, 2016-11-02]
EMA 1.07912 0.0331647 0.242043 0.285714 [2009-08-03, 2010-03-25]
RSI 1.03938 0.0305914 0.219782 0.346154 [2010-11-08, 2014-04-29]
BBI 0.858419 0.0208905 0.165719 0.334951 [2015-05-26, 2018-12-05]
EMA1 0.848388 0.0194674 0.149806 0.25641 [2009-08-03, 2014-07-10]
None 0.649806 0.000651047 0.0519684 0 [2015-06-08, 2016-01-28]
MA1 0.362528 -0.0133998 -0.0457156 0.405405 [2015-06-08, 2018-11-26]
PSY 0.335471 -0.015234 -0.0597818 0.545455 [2010-03-03, 2012-01-05]
CCI 0.297372 -0.0183251 -0.078109 0.405063 [2015-06-08, 2018-12-05]
MACD -0.248712 -0.075951 -0.368951 0.679245 [2009-08-03, 2018-08-06]
#11 回测的4个主要指标,包括总回报率、最大回撤夏普率和波动
# get_eval4_bar(self, sort_by=[])
pa.get_eval4_bar()

在上证50上的择时表现¶

#2 设定回测策略 id 
# 注意!注意!注意!这里的id是在 我的策略里面的编译运行的algorithmId,在浏览器地址里面复制一下
pa = parameter_analysis('5473a8b9e878d36100307a7d1c0405ee')
#3 运行回测
pa.get_backtest_data(file_name = 'results1.pkl',
                          running_max = 10,
                          benchmark_id = None,
                          start_date = '2008-12-05',
                           end_date = '2018-12-05',
                          frequency = 'day',
                          initial_cash = '100000000',
                          param_names = ['stock','tech'],
                          param_values = [['000016.XSHG'],['None','MA','MA1','EMA','EMA1','MACD','KDJ','CCI','RSI','BOLL','BIAS','TRIX','BBI','PSY']]
                          )
【已完成|运行中|待运行】: [0|0|14]. [0|10|4]. [0|10|4]. [1|9|4]. [1|10|3]. [5|6|3]. [11|3|0]. [11|3|0]. 
【回测完成】总用时:287秒(即0.08小时)。
#数据读取
pa.read_backtest_data('results1.pkl')
#查看回测结果指标
df1 = pa.evaluations_df
df1
.dataframe thead tr:only-child th { text-align: right; } .dataframe thead th { text-align: left; } .dataframe tbody tr th { vertical-align: top; }
stock tech __version algorithm_return algorithm_volatility alpha annual_algo_return annual_bm_return avg_position_days avg_trade_return ... max_drawdown_period max_leverage period_label profit_loss_ratio sharpe sortino trading_days treasury_return win_count win_ratio
0 000016.XSHG None 101 0.642979 0.250999 0.000786761 0.0523422 0.0515543 2433 0 ... [2009-08-03, 2014-03-20] 0 2018-12 0 0.0491725 0.0656579 2433 0.400219 0 0
1 000016.XSHG MA 101 0.739057 0.164764 0.0135366 0.0585056 0.0515543 1225 0.00678877 ... [2009-08-03, 2010-09-09] 0 2018-12 1.38853 0.112316 0.156047 2433 0.400219 31 0.236641
2 000016.XSHG MA1 101 0.08402 0.16995 -0.0369882 0.00832426 0.0515543 1203 0.00501458 ... [2009-08-03, 2014-11-06] 0 2018-12 1.07069 -0.186382 -0.228336 2433 0.400219 26 0.351351
3 000016.XSHG EMA 101 0.408567 0.16615 -0.00916891 0.0358275 0.0515543 1242 0.00482895 ... [2009-08-03, 2014-07-10] 0 2018-12 1.28927 -0.0251127 -0.0346964 2433 0.400219 35 0.263158
4 000016.XSHG EMA1 101 0.381043 0.175459 -0.011881 0.0337293 0.0515543 1273 0.0131946 ... [2009-08-03, 2014-07-10] 0 2018-12 1.33819 -0.0357391 -0.0477358 2433 0.400219 12 0.272727
5 000016.XSHG MACD 101 -0.0861159 0.18631 -0.0555538 -0.00921046 0.0515543 1185 0.00108833 ... [2009-08-03, 2014-03-10] 0 2018-12 0.981093 -0.264132 -0.266894 2433 0.400219 72 0.679245
6 000016.XSHG KDJ 101 0.609182 0.185087 0.00385646 0.0500971 0.0515543 1186 0.00339378 ... [2015-06-05, 2015-08-25] 0 2018-12 1.22866 0.0545535 0.0700506 2433 0.400219 154 0.681416
7 000016.XSHG CCI 101 0.434146 0.171199 -0.00763085 0.0377448 0.0515543 1240 0.00808786 ... [2009-11-18, 2014-07-11] 0 2018-12 1.22096 -0.0131728 -0.0174062 2433 0.400219 30 0.375
8 000016.XSHG RSI 101 0.757592 0.168756 0.0144082 0.0596593 0.0515543 1259 0.0369311 ... [2009-08-03, 2014-07-11] 0 2018-12 1.69349 0.116495 0.160392 2433 0.400219 7 0.259259
9 000016.XSHG BOLL 101 0.767735 0.135782 0.0168802 0.060286 0.0515543 783 0.0172371 ... [2015-04-27, 2018-10-11] 0 2018-12 1.54801 0.149402 0.215418 2433 0.400219 16 0.347826
10 000016.XSHG BIAS 101 0.640877 0.169366 0.00689045 0.0522038 0.0515543 1233 0.0106898 ... [2015-06-08, 2016-01-05] 0 2018-12 1.25445 0.0720561 0.0968489 2433 0.400219 31 0.418919
11 000016.XSHG TRIX 101 0.640877 0.169366 0.00689045 0.0522038 0.0515543 1233 0.0106898 ... [2015-06-08, 2016-01-05] 0 2018-12 1.25445 0.0720561 0.0968489 2433 0.400219 31 0.418919
12 000016.XSHG BBI 101 0.534281 0.164124 6.30206e-05 0.0449667 0.0515543 1209 0.00368511 ... [2009-10-26, 2012-09-18] 0 2018-12 1.23422 0.0302619 0.042997 2433 0.400219 66 0.30137
13 000016.XSHG PSY 101 -0.0858294 0.182729 -0.055268 -0.00917855 0.0515543 1261 0.00110837 ... [2009-05-13, 2016-01-28] 0 2018-12 0.943645 -0.269134 -0.333688 2433 0.400219 19 0.558824

14 rows × 28 columns

#7 回报率折线图    
pa.plot_returns()
#指标回测收益列表
df1.index = df1['tech'].values
del df1['tech']
df1 = df1[['algorithm_return','alpha','sharpe','win_ratio','max_drawdown_period']]
df1.sort_values('algorithm_return',ascending=0)
.dataframe thead tr:only-child th { text-align: right; } .dataframe thead th { text-align: left; } .dataframe tbody tr th { vertical-align: top; }
algorithm_return alpha sharpe win_ratio max_drawdown_period
BOLL 0.767735 0.0168802 0.149402 0.347826 [2015-04-27, 2018-10-11]
RSI 0.757592 0.0144082 0.116495 0.259259 [2009-08-03, 2014-07-11]
MA 0.739057 0.0135366 0.112316 0.236641 [2009-08-03, 2010-09-09]
None 0.642979 0.000786761 0.0491725 0 [2009-08-03, 2014-03-20]
BIAS 0.640877 0.00689045 0.0720561 0.418919 [2015-06-08, 2016-01-05]
TRIX 0.640877 0.00689045 0.0720561 0.418919 [2015-06-08, 2016-01-05]
KDJ 0.609182 0.00385646 0.0545535 0.681416 [2015-06-05, 2015-08-25]
BBI 0.534281 6.30206e-05 0.0302619 0.30137 [2009-10-26, 2012-09-18]
CCI 0.434146 -0.00763085 -0.0131728 0.375 [2009-11-18, 2014-07-11]
EMA 0.408567 -0.00916891 -0.0251127 0.263158 [2009-08-03, 2014-07-10]
EMA1 0.381043 -0.011881 -0.0357391 0.272727 [2009-08-03, 2014-07-10]
MA1 0.08402 -0.0369882 -0.186382 0.351351 [2009-08-03, 2014-11-06]
PSY -0.0858294 -0.055268 -0.269134 0.558824 [2009-05-13, 2016-01-28]
MACD -0.0861159 -0.0555538 -0.264132 0.679245 [2009-08-03, 2014-03-10]
#2 设定回测策略 id 
# 注意!注意!注意!这里的id是在 我的策略里面的编译运行的algorithmId,在浏览器地址里面复制一下
pa = parameter_analysis('5473a8b9e878d36100307a7d1c0405ee')
#3 运行回测
pa.get_backtest_data(file_name = 'results1.pkl',
                          running_max = 10,
                          benchmark_id = None,
                          start_date = '2008-12-05',
                           end_date = '2018-12-05',
                          frequency = 'day',
                          initial_cash = '100000000',
                          param_names = ['stock','tech'],
                          param_values = [['000905.XSHG'],['None','MA','MA1','EMA','EMA1','MACD','KDJ','CCI','RSI','BOLL','BIAS','TRIX','BBI','PSY']]
                          )
【已完成|运行中|待运行】: [0|0|14]. [0|10|4]. [0|10|4]. [9|1|4]. [已用0.031时,尚余0.012时,请不要关闭浏览器]. [10|4|0]. [10|4|0]. 
【回测完成】总用时:207秒(即0.06小时)。
#数据读取
pa.read_backtest_data('results1.pkl')
#查看回测结果指标
df2 = pa.evaluations_df
df2
.dataframe thead tr:only-child th { text-align: right; } .dataframe thead th { text-align: left; } .dataframe tbody tr th { vertical-align: top; }
stock tech __version algorithm_return algorithm_volatility alpha annual_algo_return annual_bm_return avg_position_days avg_trade_return ... max_drawdown_period max_leverage period_label profit_loss_ratio sharpe sortino trading_days treasury_return win_count win_ratio
0 000905.XSHG None 101 1.2215 0.282976 0.0364774 0.0854734 0.0515543 2433 0 ... [2015-06-12, 2018-10-18] 0 2018-12 0 0.160697 0.193085 2433 0.400219 0 0
1 000905.XSHG MA 101 1.52321 0.188519 0.0555421 0.099771 0.0515543 1362 0.0101608 ... [2015-06-12, 2018-12-05] 0 2018-12 1.43893 0.317056 0.37585 2433 0.400219 36 0.292683
2 000905.XSHG MA1 101 0.936711 0.200405 0.0255997 0.070279 0.0515543 1365 0.013663 ... [2015-06-12, 2018-11-26] 0 2018-12 1.25068 0.151089 0.176681 2433 0.400219 28 0.388889
3 000905.XSHG EMA 101 1.81217 0.186937 0.0679784 0.112092 0.0515543 1370 0.0134375 ... [2015-06-12, 2018-12-05] 0 2018-12 1.55633 0.38565 0.459733 2433 0.400219 32 0.293578
4 000905.XSHG EMA1 101 1.13677 0.198083 0.0367256 0.0811446 0.0515543 1378 0.0336433 ... [2015-06-12, 2018-11-26] 0 2018-12 1.52258 0.207714 0.243185 2433 0.400219 10 0.243902
5 000905.XSHG MACD 101 -0.0722797 0.206934 -0.0525063 -0.00767947 0.0515543 1152 0.00176669 ... [2015-06-17, 2018-10-18] 0 2018-12 0.995147 -0.230409 -0.278218 2433 0.400219 76 0.697248
6 000905.XSHG KDJ 101 1.06872 0.216676 0.0325344 0.0775551 0.0515543 1129 0.0050922 ... [2015-06-11, 2018-10-18] 0 2018-12 1.2068 0.173324 0.215837 2433 0.400219 159 0.719457
7 000905.XSHG CCI 101 1.05657 0.188599 0.0326518 0.076903 0.0515543 1315 0.0135211 ... [2015-06-12, 2018-10-08] 0 2018-12 1.35622 0.195669 0.228403 2433 0.400219 30 0.410959
8 000905.XSHG RSI 101 1.43529 0.194841 0.0514413 0.0957704 0.0515543 1311 0.0650183 ... [2015-06-12, 2018-11-29] 0 2018-12 1.79484 0.286236 0.337042 2433 0.400219 9 0.346154
9 000905.XSHG BOLL 101 0.493288 0.134002 -7.46938e-05 0.0420628 0.0515543 580 0.0150994 ... [2009-11-23, 2013-01-25] 0 2018-12 1.74083 0.0153941 0.0179303 2433 0.400219 15 0.441176
10 000905.XSHG BIAS 101 1.16579 0.190651 0.038455 0.0826447 0.0515543 1213 0.0148088 ... [2015-08-17, 2018-10-12] 0 2018-12 1.35992 0.223679 0.273367 2433 0.400219 37 0.506849
11 000905.XSHG TRIX 101 1.16579 0.190651 0.038455 0.0826447 0.0515543 1213 0.0148088 ... [2015-08-17, 2018-10-12] 0 2018-12 1.35992 0.223679 0.273367 2433 0.400219 37 0.506849
12 000905.XSHG BBI 101 1.80322 0.18043 0.0677715 0.111728 0.0515543 1356 0.00674405 ... [2015-06-12, 2018-12-05] 0 2018-12 1.45928 0.397539 0.48107 2433 0.400219 69 0.338235
13 000905.XSHG PSY 101 -0.121383 0.156835 -0.0558555 -0.0132089 0.0515543 699 -0.00108227 ... [2015-08-17, 2018-10-18] 0 2018-12 0.882128 -0.339266 -0.395475 2433 0.400219 16 0.615385

14 rows × 28 columns

#7 回报率折线图    
pa.plot_returns()
#指标回测收益列表
df2.index = df2['tech'].values
del df2['tech']
df2 = df2[['algorithm_return','alpha','sharpe','win_ratio','max_drawdown_period']]
df2.sort_values('algorithm_return',ascending=0)
.dataframe thead tr:only-child th { text-align: right; } .dataframe thead th { text-align: left; } .dataframe tbody tr th { vertical-align: top; }
algorithm_return alpha sharpe win_ratio max_drawdown_period
EMA 1.81217 0.0679784 0.38565 0.293578 [2015-06-12, 2018-12-05]
BBI 1.80322 0.0677715 0.397539 0.338235 [2015-06-12, 2018-12-05]
MA 1.52321 0.0555421 0.317056 0.292683 [2015-06-12, 2018-12-05]
RSI 1.43529 0.0514413 0.286236 0.346154 [2015-06-12, 2018-11-29]
None 1.2215 0.0364774 0.160697 0 [2015-06-12, 2018-10-18]
BIAS 1.16579 0.038455 0.223679 0.506849 [2015-08-17, 2018-10-12]
TRIX 1.16579 0.038455 0.223679 0.506849 [2015-08-17, 2018-10-12]
EMA1 1.13677 0.0367256 0.207714 0.243902 [2015-06-12, 2018-11-26]
KDJ 1.06872 0.0325344 0.173324 0.719457 [2015-06-11, 2018-10-18]
CCI 1.05657 0.0326518 0.195669 0.410959 [2015-06-12, 2018-10-08]
MA1 0.936711 0.0255997 0.151089 0.388889 [2015-06-12, 2018-11-26]
BOLL 0.493288 -7.46938e-05 0.0153941 0.441176 [2009-11-23, 2013-01-25]
MACD -0.0722797 -0.0525063 -0.230409 0.697248 [2015-06-17, 2018-10-18]
PSY -0.121383 -0.0558555 -0.339266 0.615385 [2015-08-17, 2018-10-18]
#设定回测策略 id 
# 注意!注意!注意!这里的id是在 我的策略里面的编译运行的algorithmId,在浏览器地址里面复制一下
pa = parameter_analysis('5473a8b9e878d36100307a7d1c0405ee')
#3 运行回测
pa.get_backtest_data(file_name = 'results3.pkl',
                          running_max = 10,
                          benchmark_id = None,
                          start_date = '2008-12-05',
                           end_date = '2018-12-05',
                          frequency = 'day',
                          initial_cash = '100000000',
                          param_names = ['stock','tech'],
                          param_values = [['000852.XSHG'],['None','MA','MA1','EMA','EMA1','MACD','KDJ','CCI','RSI','BOLL','BIAS','TRIX','BBI','PSY']]
                          )
【已完成|运行中|待运行】: [0|0|14]. [0|10|4]. [0|10|4]. [9|1|4]. [已用0.031时,尚余0.012时,请不要关闭浏览器]. [10|4|0]. [10|4|0]. 
【回测完成】总用时:207秒(即0.06小时)。
#数据读取
pa.read_backtest_data('results3.pkl')
#查看回测结果指标
df3 = pa.evaluations_df
df3
.dataframe thead tr:only-child th { text-align: right; } .dataframe thead th { text-align: left; } .dataframe tbody tr th { vertical-align: top; }
stock tech __version algorithm_return algorithm_volatility alpha annual_algo_return annual_bm_return avg_position_days avg_trade_return ... max_drawdown_period max_leverage period_label profit_loss_ratio sharpe sortino trading_days treasury_return win_count win_ratio
0 000852.XSHG None 101 -0.205192 0.205477 -0.0666424 -0.0233216 0.0515543 1011 0 ... [2015-06-12, 2018-10-18] 0 2018-12 0 -0.308169 -0.226447 2433 0.400219 0 0
1 000852.XSHG MA 101 0.457119 0.128087 -0.00196864 0.0394407 0.0515543 506 0.0176021 ... [2015-06-12, 2018-09-27] 0 2018-12 1.33887 -0.00436647 -0.00337005 2433 0.400219 13 0.276596
2 000852.XSHG MA1 101 0.0670167 0.138743 -0.0350068 0.00668756 0.0515543 518 0.0216535 ... [2015-06-12, 2018-11-09] 0 2018-12 1.05928 -0.240102 -0.177645 2433 0.400219 9 0.346154
3 000852.XSHG EMA 101 0.260259 0.123243 -0.0172183 0.0240535 0.0515543 477 0.0127481 ... [2015-06-12, 2018-11-30] 0 2018-12 1.28325 -0.129391 -0.0990981 2433 0.400219 10 0.212766
4 000852.XSHG EMA1 101 0.260449 0.125677 -0.017087 0.0240693 0.0515543 478 0.0228083 ... [2015-06-12, 2018-11-29] 0 2018-12 1.54503 -0.126759 -0.0971304 2433 0.400219 5 0.294118
5 000852.XSHG MACD 101 -0.476457 0.153105 -0.106142 -0.0643331 0.0515543 450 -0.0115438 ... [2015-06-12, 2018-10-18] 0 2018-12 0.646398 -0.681446 -0.389946 2433 0.400219 24 0.585366
6 000852.XSHG KDJ 101 -0.33013 0.164726 -0.0824113 -0.0403346 0.0515543 492 -0.00143382 ... [2015-06-12, 2018-10-18] 0 2018-12 0.867576 -0.487686 -0.344572 2433 0.400219 59 0.655556
7 000852.XSHG CCI 101 0.566856 0.135013 0.00563478 0.047225 0.0515543 519 0.0271793 ... [2015-06-12, 2015-09-25] 0 2018-12 1.4674 0.053513 0.0411482 2433 0.400219 11 0.366667
8 000852.XSHG RSI 101 0.384848 0.121748 -0.00710921 0.0340216 0.0515543 422 0.0473898 ... [2015-06-12, 2018-11-29] 0 2018-12 1.77654 -0.0491049 -0.038269 2433 0.400219 4 0.333333
9 000852.XSHG BOLL 101 0.909757 0.0903433 0.0280851 0.0687388 0.0515543 208 0.0862449 ... [2015-11-25, 2018-11-26] 0 2018-12 4.09809 0.318107 0.2916 2433 0.400219 3 0.272727
10 000852.XSHG BIAS 101 0.24299 0.143016 -0.0191681 0.0226026 0.0515543 522 0.013734 ... [2015-11-25, 2018-10-12] 0 2018-12 1.18594 -0.121646 -0.0912827 2433 0.400219 14 0.451613
11 000852.XSHG TRIX 101 0.24299 0.143016 -0.0191681 0.0226026 0.0515543 522 0.013734 ... [2015-11-25, 2018-10-12] 0 2018-12 1.18594 -0.121646 -0.0912827 2433 0.400219 14 0.451613
12 000852.XSHG BBI 101 0.219299 0.125023 -0.0207981 0.0205826 0.0515543 503 0.00469684 ... [2015-11-25, 2018-11-30] 0 2018-12 1.19412 -0.15531 -0.117911 2433 0.400219 26 0.305882
13 000852.XSHG PSY 101 -0.307994 0.132071 -0.0784478 -0.0371233 0.0515543 368 -0.0222212 ... [2015-08-17, 2018-10-18] 0 2018-12 0.585518 -0.583953 -0.43143 2433 0.400219 7 0.538462

14 rows × 28 columns

#7 回报率折线图    
pa.plot_returns()
#指标回测收益列表
df3.index = df3['tech'].values
del df3['tech']
df3 = df3[['algorithm_return','alpha','sharpe','win_ratio','max_drawdown_period']]
df3.sort_values('algorithm_return',ascending=0)
.dataframe thead tr:only-child th { text-align: right; } .dataframe thead th { text-align: left; } .dataframe tbody tr th { vertical-align: top; }
algorithm_return alpha sharpe win_ratio max_drawdown_period
BOLL 0.909757 0.0280851 0.318107 0.272727 [2015-11-25, 2018-11-26]
CCI 0.566856 0.00563478 0.053513 0.366667 [2015-06-12, 2015-09-25]
MA 0.457119 -0.00196864 -0.00436647 0.276596 [2015-06-12, 2018-09-27]
RSI 0.384848 -0.00710921 -0.0491049 0.333333 [2015-06-12, 2018-11-29]
EMA1 0.260449 -0.017087 -0.126759 0.294118 [2015-06-12, 2018-11-29]
EMA 0.260259 -0.0172183 -0.129391 0.212766 [2015-06-12, 2018-11-30]
BIAS 0.24299 -0.0191681 -0.121646 0.451613 [2015-11-25, 2018-10-12]
TRIX 0.24299 -0.0191681 -0.121646 0.451613 [2015-11-25, 2018-10-12]
BBI 0.219299 -0.0207981 -0.15531 0.305882 [2015-11-25, 2018-11-30]
MA1 0.0670167 -0.0350068 -0.240102 0.346154 [2015-06-12, 2018-11-09]
None -0.205192 -0.0666424 -0.308169 0 [2015-06-12, 2018-10-18]
PSY -0.307994 -0.0784478 -0.583953 0.538462 [2015-08-17, 2018-10-18]
KDJ -0.33013 -0.0824113 -0.487686 0.655556 [2015-06-12, 2018-10-18]
MACD -0.476457 -0.106142 -0.681446 0.585366 [2015-06-12, 2018-10-18]

参数加强¶

KDJ沪深300加强¶

#设定回测策略 id 
# 注意!注意!注意!这里的id是在 我的策略里面的编译运行的algorithmId,在浏览器地址里面复制一下
pa = parameter_analysis('5473a8b9e878d36100307a7d1c0405ee')
#运行回测
pa.get_backtest_data(file_name = 'results4.pkl',
                          running_max = 10,
                          benchmark_id = None,
                          start_date = '2008-12-05',
                           end_date = '2018-12-05',
                          frequency = 'day',
                          initial_cash = '100000000',
                          param_names = ['para1','para2'],
                          param_values = [[5,7,9,11,13,15,17,19,36],[3,4]]
                          )
【已完成|运行中|待运行】: [0|0|18]. [0|10|8]. [0|10|8]. [3|7|8]. [已用0.031时,尚余0.025时,请不要关闭浏览器]. [10|3|5]. [已用0.041时,尚余0.033时,请不要关闭浏览器]. [10|8|0]. [10|8|0]. [10|8|0]. [11|7|0]. [13|5|0]. [15|3|0]. [16|2|0]. [16|2|0]. [17|1|0]. [17|1|0]. 
【回测完成】总用时:495秒(即0.14小时)。
#数据读取
pa.read_backtest_data('results4.pkl')
#查看回测结果指标
df4 = pa.evaluations_df
df4
.dataframe thead tr:only-child th { text-align: right; } .dataframe thead th { text-align: left; } .dataframe tbody tr th { vertical-align: top; }
para1 para2 __version algorithm_return algorithm_volatility alpha annual_algo_return annual_bm_return avg_position_days avg_trade_return ... max_drawdown_period max_leverage period_label profit_loss_ratio sharpe sortino trading_days treasury_return win_count win_ratio
0 5 3 101 1.32556 0.188127 0.0435398 0.0905915 0.052146 1162 0.00435751 ... [2015-06-12, 2015-08-26] 0 2018-12 1.30938 0.268923 0.346242 2433 0.400219 178 0.676806
1 5 4 101 1.51813 0.186704 0.0525915 0.0995434 0.052146 1163 0.00512992 ... [2015-06-03, 2015-08-26] 0 2018-12 1.35805 0.318918 0.41437 2433 0.400219 167 0.701681
2 7 3 101 1.5104 0.185536 0.052299 0.0991958 0.052146 1163 0.00527449 ... [2018-01-26, 2018-10-18] 0 2018-12 1.36954 0.319052 0.416883 2433 0.400219 162 0.704348
3 7 4 101 1.06277 0.186579 0.0303181 0.0772365 0.052146 1158 0.00487555 ... [2015-06-02, 2015-08-26] 0 2018-12 1.28123 0.199575 0.257343 2433 0.400219 151 0.70892
4 9 3 101 1.19036 0.187656 0.0368768 0.0839002 0.052146 1176 0.00491715 ... [2015-06-03, 2015-08-26] 0 2018-12 1.3062 0.233939 0.297756 2433 0.400219 161 0.715556
5 9 4 101 0.90584 0.185912 0.0216127 0.0685133 0.052146 1157 0.00468831 ... [2015-06-02, 2015-08-26] 0 2018-12 1.28237 0.15337 0.1937 2433 0.400219 145 0.717822
6 11 3 101 0.570858 0.189836 0.000310575 0.0474994 0.052146 NaN NaN ... [2015-06-03, 2018-10-18] 0 2018-12 NaN 0.0395049 0.0497573 2433 0.400219 NaN NaN
7 11 4 101 0.374371 0.190401 -0.0140119 0.033215 0.052146 1183 0.00310097 ... [2009-08-04, 2014-03-10] 0 2018-12 1.17734 -0.0356352 -0.0425721 2433 0.400219 136 0.701031
8 13 3 101 0.216007 0.188325 -0.0267691 0.0202991 0.052146 1166 0.00239633 ... [2015-06-03, 2018-10-18] 0 2018-12 1.11778 -0.104611 -0.128373 2433 0.400219 151 0.70892
9 13 4 101 0.142828 0.189085 -0.0333495 0.0138127 0.052146 1172 0.00219321 ... [2009-08-04, 2014-03-10] 0 2018-12 1.09784 -0.138495 -0.166075 2433 0.400219 128 0.691892
10 15 3 101 0.43143 0.187181 -0.00945798 0.0375427 0.052146 1163 0.0031426 ... [2009-08-04, 2014-03-10] 0 2018-12 1.1774 -0.0131282 -0.0166523 2433 0.400219 144 0.692308
11 15 4 101 0.656516 0.187634 0.00617672 0.0532299 0.052146 1179 0.00420369 ... [2015-06-03, 2016-01-28] 0 2018-12 1.2248 0.0705091 0.0892412 2433 0.400219 133 0.71123
12 17 3 101 0.713205 0.190578 0.00960509 0.0568778 0.052146 1174 0.00399475 ... [2009-08-04, 2014-03-10] 0 2018-12 1.24239 0.0885614 0.112178 2433 0.400219 150 0.714286
13 17 4 101 0.635459 0.189647 0.0046534 0.0518463 0.052146 1180 0.00429904 ... [2010-11-11, 2014-03-10] 0 2018-12 1.21608 0.0624651 0.0792159 2433 0.400219 133 0.726776
14 19 3 101 0.91315 0.189832 0.0217144 0.0689338 0.052146 1158 0.00458051 ... [2015-06-03, 2018-10-18] 0 2018-12 1.27782 0.152418 0.194284 2433 0.400219 148 0.718447
15 19 4 101 0.743617 0.188394 0.0116509 0.0587905 0.052146 1167 0.00476142 ... [2015-06-03, 2018-10-18] 0 2018-12 1.24401 0.0997401 0.126731 2433 0.400219 130 0.726257
16 36 3 101 0.729007 0.190386 0.010605 0.0578754 0.052146 1176 0.00449526 ... [2015-06-02, 2016-01-28] 0 2018-12 1.26591 0.09389 0.119854 2433 0.400219 131 0.715847
17 36 4 101 0.724978 0.189918 0.0104203 0.0576219 0.052146 1183 0.00491793 ... [2009-08-04, 2013-06-28] 0 2018-12 1.2626 0.0927869 0.118362 2433 0.400219 122 0.721893

18 rows × 28 columns

#7 回报率折线图    
pa.plot_returns()
#指标回测收益列表
df4 = df4[['para1','para2','algorithm_return','alpha','sharpe','win_ratio','max_drawdown_period']]
df4.sort_values('algorithm_return',ascending=0)
.dataframe thead tr:only-child th { text-align: right; } .dataframe thead th { text-align: left; } .dataframe tbody tr th { vertical-align: top; }
para1 para2 algorithm_return alpha sharpe win_ratio max_drawdown_period
1 5 4 1.51813 0.0525915 0.318918 0.701681 [2015-06-03, 2015-08-26]
2 7 3 1.5104 0.052299 0.319052 0.704348 [2018-01-26, 2018-10-18]
0 5 3 1.32556 0.0435398 0.268923 0.676806 [2015-06-12, 2015-08-26]
4 9 3 1.19036 0.0368768 0.233939 0.715556 [2015-06-03, 2015-08-26]
3 7 4 1.06277 0.0303181 0.199575 0.70892 [2015-06-02, 2015-08-26]
14 19 3 0.91315 0.0217144 0.152418 0.718447 [2015-06-03, 2018-10-18]
5 9 4 0.90584 0.0216127 0.15337 0.717822 [2015-06-02, 2015-08-26]
15 19 4 0.743617 0.0116509 0.0997401 0.726257 [2015-06-03, 2018-10-18]
16 36 3 0.729007 0.010605 0.09389 0.715847 [2015-06-02, 2016-01-28]
17 36 4 0.724978 0.0104203 0.0927869 0.721893 [2009-08-04, 2013-06-28]
12 17 3 0.713205 0.00960509 0.0885614 0.714286 [2009-08-04, 2014-03-10]
11 15 4 0.656516 0.00617672 0.0705091 0.71123 [2015-06-03, 2016-01-28]
13 17 4 0.635459 0.0046534 0.0624651 0.726776 [2010-11-11, 2014-03-10]
6 11 3 0.570858 0.000310575 0.0395049 NaN [2015-06-03, 2018-10-18]
10 15 3 0.43143 -0.00945798 -0.0131282 0.692308 [2009-08-04, 2014-03-10]
7 11 4 0.374371 -0.0140119 -0.0356352 0.701031 [2009-08-04, 2014-03-10]
8 13 3 0.216007 -0.0267691 -0.104611 0.70892 [2015-06-03, 2018-10-18]
9 13 4 0.142828 -0.0333495 -0.138495 0.691892 [2009-08-04, 2014-03-10]

BOLL中证1000参数测试¶

#设定回测策略 id 
# 注意!注意!注意!这里的id是在 我的策略里面的编译运行的algorithmId,在浏览器地址里面复制一下
pa = parameter_analysis('f0c6eb90370078648d793bf93bb876b0')
#运行回测
pa.get_backtest_data(file_name = 'results6.pkl',
                          running_max = 10,
                          benchmark_id = None,
                          start_date = '2008-12-05',
                           end_date = '2018-12-05',
                          frequency = 'day',
                          initial_cash = '100000000',
                          param_names = ['para1'],
                          param_values = [['10','11','12','13','14','15','16','17','18','19','20','21','22','23','24','25','26','27','28','30']]
                          )
 【已完成|运行中|待运行】: [0|0|20]. [0|10|10]. [0|10|10]. [10|0|10]. [已用0.032时,尚余0.032时,请不要关闭浏览器]. [10|10|0]. [11|9|0]. 
【回测完成】总用时:217秒(即0.06小时)。
 
#数据读取
pa.read_backtest_data('results6.pkl')
#查看回测结果指标
df6 = pa.evaluations_df
df6
.dataframe thead tr:only-child th { text-align: right; } .dataframe thead th { text-align: left; } .dataframe tbody tr th { vertical-align: top; }
para1 __version algorithm_return algorithm_volatility alpha annual_algo_return annual_bm_return avg_position_days avg_trade_return benchmark_return ... max_drawdown_period max_leverage period_label profit_loss_ratio sharpe sortino trading_days treasury_return win_count win_ratio
0 10 101 0.865643 0.0894054 0.0378796 0.0661754 -0.0232902 219 0.0375561 -0.204943 ... [2015-06-12, 2015-11-30] 0 2018-12 3.38193 0.292772 0.251191 2433 0.400219 11 0.55
1 11 101 0.58006 0.0864753 0.0193872 0.0481284 -0.0232902 193 0.0321747 -0.204943 ... [2015-10-20, 2018-03-20] 0 2018-12 2.43668 0.0939964 0.0788521 2433 0.400219 9 0.5
2 12 101 0.493471 0.0894678 0.0141437 0.042076 -0.0232902 216 0.0278542 -0.204943 ... [2015-10-27, 2018-07-31] 0 2018-12 2.03417 0.0232039 0.0194102 2433 0.400219 9 0.473684
3 13 101 0.818732 0.0906441 0.0358485 0.0633891 -0.0232902 251 0.0380457 -0.204943 ... [2016-04-15, 2017-03-23] 0 2018-12 4.33628 0.258033 0.225088 2433 0.400219 10 0.526316
4 14 101 0.51269 0.0957402 0.0169971 0.043446 -0.0232902 257 0.0279937 -0.204943 ... [2015-11-25, 2017-03-24] 0 2018-12 2.29674 0.0359934 0.028992 2433 0.400219 10 0.526316
5 15 101 0.402707 0.0951702 0.00871467 0.0353839 -0.0232902 257 0.0229583 -0.204943 ... [2015-11-25, 2017-08-25] 0 2018-12 1.8705 -0.0485035 -0.0389315 2433 0.400219 9 0.45
6 16 101 0.743311 0.0987706 0.0331036 0.0587714 -0.0232902 253 0.0613841 -0.204943 ... [2015-11-25, 2016-07-01] 0 2018-12 2.62719 0.19005 0.15913 2433 0.400219 8 0.533333
7 17 101 0.843695 0.102206 0.0403671 0.0648797 -0.0232902 264 0.0677565 -0.204943 ... [2015-11-25, 2016-07-01] 0 2018-12 2.77774 0.243427 0.206631 2433 0.400219 6 0.428571
8 18 101 0.741596 0.101487 0.033928 0.0586643 -0.0232902 256 0.0637892 -0.204943 ... [2015-11-25, 2018-11-26] 0 2018-12 2.36098 0.183909 0.155369 2433 0.400219 5 0.357143
9 19 101 0.965675 0.096355 0.0459119 0.0719128 -0.0232902 249 0.0769897 -0.204943 ... [2015-11-25, 2018-11-26] 0 2018-12 4.29575 0.3312 0.296015 2433 0.400219 4 0.307692
10 20 101 0.909757 0.0903433 0.0410385 0.0687388 -0.0232902 208 0.0862449 -0.204943 ... [2015-11-25, 2018-11-26] 0 2018-12 4.09809 0.318107 0.2916 2433 0.400219 3 0.272727
11 21 101 1.01866 0.0908123 0.0472858 0.0748465 -0.0232902 214 0.108241 -0.204943 ... [2015-11-25, 2018-11-26] 0 2018-12 4.25828 0.38372 0.353497 2433 0.400219 2 0.2
12 22 101 0.963169 0.0926881 0.0447195 0.0717723 -0.0232902 219 0.105523 -0.204943 ... [2015-11-25, 2018-11-30] 0 2018-12 3.62066 0.342787 0.308921 2433 0.400219 2 0.2
13 23 101 1.05844 0.0913043 0.0495816 0.0770036 -0.0232902 217 0.122116 -0.204943 ... [2015-11-25, 2018-11-30] 0 2018-12 5.1772 0.405278 0.372345 2433 0.400219 2 0.222222
14 24 101 1.02774 0.0895473 0.0474594 0.0753423 -0.0232902 202 0.1188 -0.204943 ... [2015-11-25, 2018-11-30] 0 2018-12 5.40792 0.394678 0.36954 2433 0.400219 2 0.222222
15 25 101 0.66911 0.0984739 0.0281853 0.0540499 -0.0232902 213 0.0836229 -0.204943 ... [2015-11-25, 2018-11-30] 0 2018-12 2.3265 0.142677 0.12049 2433 0.400219 2 0.2
16 26 101 0.609643 0.0978181 0.0241008 0.050128 -0.0232902 209 0.0791121 -0.204943 ... [2015-11-25, 2018-11-30] 0 2018-12 2.13805 0.103539 0.0871196 2433 0.400219 2 0.2
17 27 101 0.649213 0.0976752 0.0266846 0.0527519 -0.0232902 205 0.0905141 -0.204943 ... [2015-11-25, 2018-11-30] 0 2018-12 2.30129 0.130554 0.109956 2433 0.400219 2 0.222222
18 28 101 0.642818 0.0982901 0.0264433 0.0523317 -0.0232902 208 0.09026 -0.204943 ... [2015-11-25, 2018-11-30] 0 2018-12 2.21495 0.125462 0.106122 2433 0.400219 2 0.222222
19 30 101 0.615735 0.0950642 0.0237083 0.0505357 -0.0232902 180 0.0933548 -0.204943 ... [2015-11-25, 2017-09-26] 0 2018-12 2.54549 0.110828 0.0940018 2433 0.400219 2 0.25

20 rows × 27 columns

#7 回报率折线图    
pa.plot_returns()
#指标回测收益列表
df6 = df6[['para1','algorithm_return','alpha','sharpe','win_ratio','max_drawdown_period']]
df6.sort_values('algorithm_return',ascending=0)
.dataframe thead tr:only-child th { text-align: right; } .dataframe thead th { text-align: left; } .dataframe tbody tr th { vertical-align: top; }
para1 algorithm_return alpha sharpe win_ratio max_drawdown_period
13 23 1.05844 0.0495816 0.405278 0.222222 [2015-11-25, 2018-11-30]
14 24 1.02774 0.0474594 0.394678 0.222222 [2015-11-25, 2018-11-30]
11 21 1.01866 0.0472858 0.38372 0.2 [2015-11-25, 2018-11-26]
9 19 0.965675 0.0459119 0.3312 0.307692 [2015-11-25, 2018-11-26]
12 22 0.963169 0.0447195 0.342787 0.2 [2015-11-25, 2018-11-30]
10 20 0.909757 0.0410385 0.318107 0.272727 [2015-11-25, 2018-11-26]
0 10 0.865643 0.0378796 0.292772 0.55 [2015-06-12, 2015-11-30]
7 17 0.843695 0.0403671 0.243427 0.428571 [2015-11-25, 2016-07-01]
3 13 0.818732 0.0358485 0.258033 0.526316 [2016-04-15, 2017-03-23]
6 16 0.743311 0.0331036 0.19005 0.533333 [2015-11-25, 2016-07-01]
8 18 0.741596 0.033928 0.183909 0.357143 [2015-11-25, 2018-11-26]
15 25 0.66911 0.0281853 0.142677 0.2 [2015-11-25, 2018-11-30]
17 27 0.649213 0.0266846 0.130554 0.222222 [2015-11-25, 2018-11-30]
18 28 0.642818 0.0264433 0.125462 0.222222 [2015-11-25, 2018-11-30]
19 30 0.615735 0.0237083 0.110828 0.25 [2015-11-25, 2017-09-26]
16 26 0.609643 0.0241008 0.103539 0.2 [2015-11-25, 2018-11-30]
1 11 0.58006 0.0193872 0.0939964 0.5 [2015-10-20, 2018-03-20]
4 14 0.51269 0.0169971 0.0359934 0.526316 [2015-11-25, 2017-03-24]
2 12 0.493471 0.0141437 0.0232039 0.473684 [2015-10-27, 2018-07-31]
5 15 0.402707 0.00871467 -0.0485035 0.45 [2015-11-25, 2017-08-25]
df6.index = df6['para1'].values
df6[['para1','algorithm_return']].plot()
<matplotlib.axes._subplots.AxesSubplot at 0x7f4fddf55a50>

EMA中证500加强¶

#设定回测策略 id 
# 注意!注意!注意!这里的id是在 我的策略里面的编译运行的algorithmId,在浏览器地址里面复制一下
pa = parameter_analysis('f50a271853f177b4d3d30b14eff6c7fe')
#运行回测
pa.get_backtest_data(file_name = 'results5.pkl',
                          running_max = 10,
                          benchmark_id = None,
                          start_date = '2008-12-05',
                           end_date = '2018-12-05',
                          frequency = 'day',
                          initial_cash = '100000000',
                          param_names = ['para1','para2'],
                          param_values = [['10','11','12','13','14','15','16','17','18','19','20','21','22','23','24','25','26','27','28','30'],[3]]
                          )
【已完成|运行中|待运行】: [0|0|20]. [0|10|10]. [3|7|10]. [已用0.022时,尚余0.022时,请不要关闭浏览器]. [10|3|7]. [12|8|0]. [13|7|0]. 
【回测完成】总用时:214秒(即0.06小时)。
#数据读取
pa.read_backtest_data('results5.pkl')
#查看回测结果指标
df5 = pa.evaluations_df
df5
.dataframe thead tr:only-child th { text-align: right; } .dataframe thead th { text-align: left; } .dataframe tbody tr th { vertical-align: top; }
para1 para2 __version algorithm_return algorithm_volatility alpha annual_algo_return annual_bm_return avg_position_days avg_trade_return ... max_drawdown_period max_leverage period_label profit_loss_ratio sharpe sortino trading_days treasury_return win_count win_ratio
0 10 3 101 1.53432 0.181471 0.0418539 0.100268 0.0848782 NaN NaN ... [2015-06-12, 2018-12-05] 0 2018-12 NaN 0.332107 0.402474 2433 0.400219 NaN NaN
1 11 3 101 1.80154 0.183105 0.0528804 0.111659 0.0848782 1361 0.00707919 ... [2015-06-12, 2018-12-05] 0 2018-12 1.39735 0.391357 0.473785 2433 0.400219 66 0.338462
2 12 3 101 1.85139 0.181915 0.0550891 0.113676 0.0848782 1361 0.00750812 ... [2015-06-12, 2018-12-05] 0 2018-12 1.43433 0.405001 0.488552 2433 0.400219 64 0.338624
3 13 3 101 1.95747 0.183435 0.0588877 0.117864 0.0848782 1366 0.00807861 ... [2015-06-12, 2018-12-05] 0 2018-12 1.49935 0.424475 0.513215 2433 0.400219 58 0.324022
4 14 3 101 1.94683 0.183724 0.0584149 0.11745 0.0848782 1361 0.008254 ... [2015-11-25, 2018-11-30] 0 2018-12 1.54831 0.421555 0.510287 2433 0.400219 54 0.313953
5 15 3 101 1.71114 0.183796 0.0488681 0.107919 0.0848782 1363 0.0079412 ... [2015-06-12, 2018-11-30] 0 2018-12 1.50483 0.369534 0.44506 2433 0.400219 49 0.291667
6 16 3 101 1.47948 0.183381 0.0388361 0.0977972 0.0848782 1366 0.00781405 ... [2015-06-12, 2018-11-30] 0 2018-12 1.44679 0.315175 0.377609 2433 0.400219 46 0.289308
7 17 3 101 1.52717 0.183432 0.0410611 0.099948 0.0848782 1368 0.00817243 ... [2015-06-12, 2018-11-30] 0 2018-12 1.45932 0.326813 0.39168 2433 0.400219 45 0.292208
8 18 3 101 1.84575 0.180868 0.055009 0.113449 0.0848782 1364 0.00892756 ... [2015-11-25, 2018-11-30] 0 2018-12 1.59759 0.406093 0.491404 2433 0.400219 46 0.298701
9 19 3 101 1.6635 0.180882 0.0474507 0.105903 0.0848782 1366 0.00846828 ... [2015-11-25, 2018-11-30] 0 2018-12 1.5479 0.36434 0.440149 2433 0.400219 45 0.294118
10 20 3 101 2.20461 0.181194 0.0685876 0.12712 0.0848782 1367 0.0105176 ... [2015-11-25, 2018-11-30] 0 2018-12 1.64558 0.480813 0.582654 2433 0.400219 39 0.272727
11 21 3 101 2.69223 0.183037 0.0847354 0.143645 0.0848782 1374 0.0121224 ... [2015-11-25, 2018-11-26] 0 2018-12 1.70167 0.566249 0.688305 2433 0.400219 38 0.279412
12 22 3 101 2.02202 0.183182 0.0614396 0.120347 0.0848782 1373 0.0109319 ... [2015-06-12, 2018-11-26] 0 2018-12 1.52003 0.438617 0.531152 2433 0.400219 36 0.270677
13 23 3 101 2.52702 0.183497 0.0793642 0.138278 0.0848782 1370 0.013731 ... [2015-11-25, 2018-12-05] 0 2018-12 1.66611 0.535581 0.651291 2433 0.400219 34 0.267717
14 24 3 101 2.0179 0.186289 0.060799 0.120189 0.0848782 1370 0.0126466 ... [2015-11-25, 2018-12-05] 0 2018-12 1.50178 0.430458 0.517689 2433 0.400219 34 0.269841
15 25 3 101 2.28683 0.187264 0.070472 0.130058 0.0848782 1368 0.0139336 ... [2015-06-12, 2018-12-05] 0 2018-12 1.58324 0.480917 0.57636 2433 0.400219 34 0.290598
16 26 3 101 2.09463 0.187018 0.0635644 0.123083 0.0848782 1364 0.0131264 ... [2015-06-12, 2018-12-05] 0 2018-12 1.52505 0.444252 0.532583 2433 0.400219 33 0.275
17 27 3 101 1.88055 0.186435 0.0554284 0.114841 0.0848782 1365 0.013034 ... [2015-06-12, 2018-12-05] 0 2018-12 1.53161 0.40143 0.4784 2433 0.400219 33 0.289474
18 28 3 101 1.79408 0.188034 0.0515226 0.111355 0.0848782 1371 0.0129666 ... [2015-06-12, 2018-12-05] 0 2018-12 1.47497 0.379479 0.452258 2433 0.400219 31 0.27193
19 30 3 101 1.81217 0.186937 0.0525923 0.112092 0.0848782 1370 0.0134375 ... [2015-06-12, 2018-12-05] 0 2018-12 1.55633 0.38565 0.459733 2433 0.400219 32 0.293578

20 rows × 28 columns

#7 回报率折线图    
pa.plot_returns()
#指标回测收益列表
df5 = df5[['para1','para2','algorithm_return','alpha','sharpe','win_ratio','max_drawdown_period']]
df5.sort_values('algorithm_return',ascending=0)
.dataframe thead tr:only-child th { text-align: right; } .dataframe thead th { text-align: left; } .dataframe tbody tr th { vertical-align: top; }
para1 para2 algorithm_return alpha sharpe win_ratio max_drawdown_period
4 21 3 2.69223 0.0847354 0.566249 0.279412 [2015-11-25, 2018-11-26]
5 23 3 2.52702 0.0793642 0.535581 0.267717 [2015-11-25, 2018-12-05]
6 25 3 2.28683 0.070472 0.480917 0.290598 [2015-06-12, 2018-12-05]
7 27 3 1.88055 0.0554284 0.40143 0.289474 [2015-06-12, 2018-12-05]
8 30 3 1.81217 0.0525923 0.38565 0.293578 [2015-06-12, 2018-12-05]
0 11 3 1.80154 0.0528804 0.391357 0.338462 [2015-06-12, 2018-12-05]
1 15 3 1.71114 0.0488681 0.369534 0.291667 [2015-06-12, 2018-11-30]
3 19 3 1.6635 0.0474507 0.36434 0.294118 [2015-11-25, 2018-11-30]
2 17 3 1.52717 0.0410611 0.326813 0.292208 [2015-06-12, 2018-11-30]
df5.index = df5['para1'].values
df5[['para1','algorithm_return']].plot()
<matplotlib.axes._subplots.AxesSubplot at 0x7f234ff7f490>

全部回复

0/140

量化课程

    移动端课程