Onepy is an event-driven algorithmic trading Python library.
更新日志：Change Log
Onepy is developed using Python 3.x and depends on:
You can install them by pip and make sure they are up-to-date：
pip install pandas
pip install TA-Lib
pip install plotly
pip install funcy
pip install OnePy_trader
pip install --upgrade OnePy_trader
OnePy安装完成后复制以下代码运行即可，可以迅速了解本框架的主要功能。 记得下载好data文件夹中的文件，设置好数据读取路径。 以Forex为例：
import matplotlib.pyplot as plt
import OnePy as op
class MyStrategy(op.StrategyBase):
# 可用参数：
# list格式： self.cash, self.position, self.margin,
# self.total, self.unre_profit
def __init__(self,marketevent):
super(MyStrategy,self).__init__(marketevent)
def prenext(self):
# print(self.unre_profit[-1])
pass
def next(self):
"""这里写主要的策略思路"""
if self.i.SMA(period=30, index=-1) > self.i.SMA(period=50,index=-1):
if self.unre_profit[-1] <= 0:
self.Buy(0.1,limit=self.pips(200), # 设置止盈为200个pips，不可为负
stop=self.pct(1), # 设置止损为成交价的1%，不可为负
trailingstop=self.pips(60)) # 设置追踪止损，盈利时触发
else:
self.Sell(0.05,price=self.pips(50), # 设置挂单，默认为第二天open价格加50点，也可为负数
limit=self.pips(200),
stop=self.pips(200),
trailingstop=self.pips(60))
if self.unre_profit[-2] > self.unre_profit[-1] and self.unre_profit[-2] > 100:
self.Exitall() # 设置浮亏浮盈大于100元且出现下降时清仓
go = op.OnePiece()
Forex = op.Forex_CSVFeed(datapath='data/EUR_USD30m.csv', # 注意设置好文件存放路径
instrument='EUR_USD',
fromdate='2016-04-01',
todate='2016-05-01')
data_list = [Forex]
portfolio = op.PortfolioBase
strategy = MyStrategy
broker = op.SimulatedBroker
go.set_backtest(data_list,[strategy],portfolio,broker,'Forex')
go.set_commission(commission=30,margin=325,mult=10**5) # 手续费为点差30pips，每手保证金为325，1pips为1/(10**5)
go.set_cash(10000) # 设置初始资金
# go.set_pricetype(‘close’) # 设置成交价格为close，若不设置，默认为open
go.set_notify() # 打印交易日志
go.sunny() # 开始启动策略
# print(go.get_tlog()) # 打印交易日志
go.get_analysis('EUR_USD') # 进行交易分析
go.plot(instrument='EUR_USD',notebook=False) # 若在Jupyter notebook中运行，可将notebook设置为True
# 简易的画图，将后面想要画的选项后面的 1 删掉即可
# go.oldplot(['un_profit','re_profit','position1','cash1','total','margin1','avg_price1'])结果：
+------------------------+
| Final_Value | 14201.0 |
| Total_return | 42.01% |
| Max_Drawdown | 13.971% |
| Duration | 289.0 |
| Sharpe_Ratio | 0.91 |
+------------------------+
+------------------------------------------------------------------+
| start | 2016-04-01 00:00:00 |
| end | 2016-04-29 21:00:00 |
| beginning_balance | 10000 |
| ending_balance | 14201.0 |
| unrealized_profit | 2735.05 |
| total_net_profit | 1465.95 |
| gross_profit | 3530.4 |
| gross_loss | -2064.45 |
| profit_factor | 1.71 |
| return_on_initial_capital | 14.66 |
| annual_return_rate | 8346.873 |
| trading_period | 0 years 0 months 28 days |
| pct_time_in_market | 754.084 |
| total_num_trades | 487 |
| num_winning_trades | 282 |
| num_losing_trades | 205 |
| num_even_trades | 0 |
| pct_profitable_trades | 57.906 |
| avg_profit_per_trade | 3.01 |
| avg_profit_per_winning_trade | 12.519 |
| avg_loss_per_losing_trade | -10.07 |
| ratio_avg_profit_win_loss | 1.243 |
| largest_profit_winning_trade | 73.5 |
| largest_loss_losing_trade | -57.0 |
| num_winning_points | 0.507 |
| num_losing_points | -0.474 |
| total_net_points | 0.032 |
| avg_points | 0.0 |
| largest_points_winning_trade | 0.011 |
| largest_points_losing_trade | -0.008 |
| avg_pct_gain_per_trade | 0.006 |
| largest_pct_winning_trade | 0.955 |
| largest_pct_losing_trade | -0.716 |
| max_consecutive_winning_trades | 38 |
| max_consecutive_losing_trades | 32 |
| avg_bars_winning_trades | 18.004 |
| avg_bars_losing_trades | 12.166 |
| max_closed_out_drawdown | -10.631 |
| max_closed_out_drawdown_start_date | 2016-04-19 18:30:00 |
| max_closed_out_drawdown_end_date | 2016-04-20 11:00:00 |
| max_closed_out_drawdown_recovery_date | 2016-04-26 09:00:00 |
| drawdown_recovery | -0.002 |
| drawdown_annualized_return | -0.001 |
| max_intra_day_drawdown | -10.751 |
| avg_yearly_closed_out_drawdown | -6.06 |
| max_yearly_closed_out_drawdown | -10.631 |
| avg_monthly_closed_out_drawdown | -1.689 |
| max_monthly_closed_out_drawdown | -9.89 |
| avg_weekly_closed_out_drawdown | -0.64 |
| max_weekly_closed_out_drawdown | -9.244 |
| avg_yearly_closed_out_runup | 11.02 |
| max_yearly_closed_out_runup | 27.051 |
| avg_monthly_closed_out_runup | 2.339 |
| max_monthly_closed_out_runup | 15.85 |
| avg_weekly_closed_out_runup | 0.816 |
| max_weekly_closed_out_runup | 11.682 |
| pct_profitable_years | 95.745 |
| best_year | 20.696 |
| worst_year | -2.93 |
| avg_year | 7.486 |
| annual_std | 4.949 |
| pct_profitable_months | 63.72 |
| best_month | 13.798 |
| worst_month | -8.54 |
| avg_month | 0.714 |
| monthly_std | 2.538 |
| pct_profitable_weeks | 57.057 |
| best_week | 11.402 |
| worst_week | -8.362 |
| avg_week | 0.184 |
| weekly_std | 1.36 |
| sharpe_ratio | 0.91 |
| sortino_ratio | 1.056 |
+------------------------------------------------------------------+
- 事件驱动回测设计 ✓
- Forex模式 ✓
- Futures模式
- Stock模式 ✓
- 多品种回测(同一模式下) ✓
- 多策略回测 ✓
- 设置手续费，保证金/手，杠杆大小 ✓
- 设置成交价格为close或者第二天open ✓
- 设置是否打印交易日志 ✓
- Plot 画图模块 ✓
- 设置Bar mode或者Tick mode
- Optimizer 参数优化模块
- To_MongoDB：自定义数据统一格式后存入数据库 ✓
- To_MongoDB：tickstory外汇数据CSV存入数据库 ✓
- To_MongoDB：tushare股票数据CSV存入数据库 ✓
- 实时采集数据存入MongoDB
- 自定义CSV数据读取 ✓
- tickstory外汇数据CSV读取 ✓
- Tushare股票数据CSV读取 ✓
- 期货数据CSV读取 ✓
- 从MongoDB数据库读取数据
- 实现做多Buy，做空Sell指令，一键平仓指令 ✓
- 按百分比pct或基点pips，挂多单(above&below)和挂空单(above&below) ✓
- 按百分比pct或基点pips，止盈止损 ✓
- 按百分比pct或基点pips，移动止损 ✓
- 自定义打印交易信息 ✓
- 技术指标Indicator模块 ✓
- OCO指令
- 挂单到时过期
- 取消挂单指令
- 暂无
- 模拟发送指令 ✓
- 模拟检查指令是否发送成功 ✓
- 打印交易日志 notify ✓
- 手续费commission，百分比类型和固定类型 ✓
- oanda接口
- 计算保证金，仓位，总利润，总额，剩余现金，收益率，全部序列化 ✓
- 输出交易记录 ✓
- 交易结果超简单分析 ✓
- 交易记录详细分析 ✓
- 结合Benchmark分析
- vnpy
- Backtrader
- PyAlgoTrade
- Zipline
- Ultra-Finance
- ProfitPy
- pybacktest
- prophet
- quant
- AlephNull
- Trading with Python
- visualize-wealth
- tia: Toolkit for integration and analysis
- QuantSoftware Toolkit
- Pinkfish
- bt
- PyThalesians
- QSTrader
- QSForex
- pysystemtrade
- QTPyLib
- RQalpha
这个回测框架内部还存在很多问题，比如交易结果分析的公式是照搬Pinkfish的，准确性还有待考证，又比如在策略中同时发出做多，做空和一键平仓信号，结果不一定为全部平仓等。这些问题还需要在接下来的应用和思考中才能发现和修改。
所以此框架主要做学习之用，若想直接拿去回测思路还请三思。
另外本人接下来一段时间要回归书本汲取新的知识了，所以OnePy更新暂时告一段落。
如果你有什么想法欢迎随时和我交流。
Wechat：chenjiayicjy，添加请注明OnePy。 感恩。
I'm very interested in your experience with Onepy.Please feel free to contact me via chenjiayicjy@126.com
Chandler_Chan