https://github.com/truthhun/parameter-optimization
参数优化(股票+期货)，基于循环遍历回测的参数优化方法
https://github.com/truthhun/parameter-optimization
Last synced: about 2 months ago
JSON representation
参数优化(股票+期货)，基于循环遍历回测的参数优化方法
- Host: GitHub
- URL: https://github.com/truthhun/parameter-optimization
- Owner: TruthHun
- Created: 2018-01-31T06:08:13.000Z (over 7 years ago)
- Default Branch: master
- Last Pushed: 2018-01-31T06:11:16.000Z (over 7 years ago)
- Last Synced: 2025-04-07T23:39:22.989Z (3 months ago)
- Language: Python
- Homepage: https://www.myquant.cn/
- Size: 1.95 KB
- Stars: 11
- Watchers: 3
- Forks: 7
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# 参数优化(股票+期货)
基于循环遍历回测的参数优化方法
## 源码
```python
# coding=utf-8
from __future__ import print_function, absolute_import, unicode_literals
import multiprocessing
import numpy as np
import pandas as pd
import talib
from gm.api import *
'''
基本思想：设定所需优化的参数数值范围及步长，将参数数值循环输入进策略，进行遍历回测，
记录每次回测结果和参数，根据某种规则将回测结果排序，找到最好的参数。
1、定义策略函数
2、多进程循环输入参数数值
3、获取回测报告，生成DataFrame格式
4、排序
本程序以双均线策略为例，优化两均线长短周期参数。
'''
# 原策略中的参数定义语句需要删除！
def init(context):
context.sec_id = 'SHSE.600000'
subscribe(symbols=context.sec_id, frequency='1d', count=31, wait_group=True)
def on_bar(context, bars):
close = context.data(symbol=context.sec_id, frequency='1d', count=31, fields='close')['close'].values
MA_short = talib.MA(close, timeperiod=context.short)
MA_long = talib.MA(close, timeperiod=context.long)
position = context.account().position(symbol=context.sec_id, side=PositionSide_Long)
if not position and not position:
if MA_short[-1] > MA_long[-1] and MA_short[-2] < MA_long[-2]:
order_target_percent(symbol=context.sec_id, percent=0.8, order_type=OrderType_Market,
position_side=PositionSide_Long)
elif position:
if MA_short[-1] < MA_long[-1] and MA_short[-2] > MA_long[-2]:
order_target_percent(symbol=context.sec_id, percent=0, order_type=OrderType_Market,
position_side=PositionSide_Long)
# 获取每次回测的报告数据
def on_backtest_finished(context, indicator):
data = [indicator['pnl_ratio'], indicator['pnl_ratio_annual'], indicator['sharp_ratio'], indicator['max_drawdown'],
context.short, context.long]
# 将回测报告加入全局list，以便记录
context.list.append(data)
def run_strategy(short, long, a_list):
from gm.model.storage import context
# 用context传入参数
context.short = short
context.long = long
# a_list一定要传入
context.list = a_list
'''
strategy_id策略ID,由系统生成
filename文件名,请与本文件名保持一致
mode实时模式:MODE_LIVE回测模式:MODE_BACKTEST
token绑定计算机的ID,可在系统设置-密钥管理中生成
backtest_start_time回测开始时间
backtest_end_time回测结束时间
backtest_adjust股票复权方式不复权:ADJUST_NONE前复权:ADJUST_PREV后复权:ADJUST_POST
backtest_initial_cash回测初始资金
backtest_commission_ratio回测佣金比例
backtest_slippage_ratio回测滑点比例
'''
run(strategy_id='strategy_id',
filename='main.py',
mode=MODE_BACKTEST,
token='token_id',
backtest_start_time='2017-05-01 08:00:00',
backtest_end_time='2017-10-01 16:00:00',
backtest_adjust=ADJUST_PREV,
backtest_initial_cash=50000,
backtest_commission_ratio=0.0001,
backtest_slippage_ratio=0.0001)
if __name__ == '__main__':
# 生成全局list
manager = multiprocessing.Manager()
a_list = manager.list()
# 循环输入参数数值回测
for short in range(5, 10, 2):
for long in range(10, 21, 5):
process = multiprocessing.Process(target=run_strategy, args=(short, long, a_list))
process.start()
process.join()
# 回测报告转化成DataFrame格式
a_list = np.array(a_list)
final = pd.DataFrame(a_list,
columns=['pnl_ratio', 'pnl_ratio_annual', 'sharp_ratio', 'max_drawdown', 'short', 'long'])
# 回测报告排序
final = final.sort_values(axis=0, ascending=False, by='pnl_ratio')
print(final)
```