akshare sma策略

import datetime
 
import pandas as pd
 
import backtrader as bt
from datetime import datetime
import matplotlib
import akshare as ak
%matplotlib inline
 
class SmaCross(bt.Strategy):
 
    # 全局设定交易策略的参数
 
    params = (('pfast', 5), ('pslow', 20),)
 
    def __init__(self):
 
          
 
        sma1 = bt.ind.SMA(period=self.p.pfast)  # fast moving average
 
        sma2 = bt.ind.SMA(period=self.p.pslow)  # slow moving average
 
        self.crossover = bt.ind.CrossOver(sma1, sma2)  # crossover signal
 
    def next(self):
 
        if self.crossover > 0:  # if fast crosses slow to the upside
 
            self.close()
 
            print(self.position)
 
            self.buy(size=1500) # enter long
 
            print("Buy {} shares".format( self.data.close[0]))
 
            print(self.position)
 
        elif self.crossover < 0:  # in the market & cross to the downside
 
            self.close()# close long position
 
            print(self.position)
 
            self.sell(size=1500)
 
            print("Sale {} shares".format(self.data.close[0]))
 
            print(self.position)
 
 
 
def bt1():
 
    # 利用 AKShare 获取股票的后复权数据，这里只获取前 6 列
    stock_hfq_df = ak.stock_zh_a_hist(symbol="000001", adjust="hfq").iloc[:, :6]
    # 处理字段命名，以符合 Backtrader 的要求
    stock_hfq_df.columns = [
    'date',
    'open',
    'close',
    'high',
    'low',
    'volume',
    ]
    # 把 date 作为日期索引，以符合 Backtrader 的要求
    stock_hfq_df.index = pd.to_datetime(stock_hfq_df['date'])
    start_date = datetime(1991, 4, 3)  # 回测开始时间
    end_date = datetime(2022, 6, 16)  # 回测结束时间
    data = bt.feeds.PandasData(dataname=stock_hfq_df,
                           fromdate=start_date,
                           todate=end_date)  # 加载数据
 
    # 初始化cerebro回测系统设置
 
    cerebro = bt.Cerebro()
 
    # 将数据传入回测系统
 
    cerebro.adddata(data)
 
    # 将交易策略加载到回测系统中
 
    cerebro.addstrategy(SmaCross)
 
    # 设置初始资本为10,000
 
    startcash = 10000
 
    cerebro.broker.setcash(startcash)
 
    # 设置交易手续费为 0.1%
 
    cerebro.broker.setcommission(commission=0.001)
 
    # 运行回测系统
 
    cerebro.run()
 
    # 获取回测结束后的总资金
 
    portvalue = cerebro.broker.getvalue()
 
    pnl = portvalue - startcash
 
    print(f'净收益: {round(pnl,2)}')
 
    # 打印结果
 
    print(f'总资金: {round(portvalue,2)}')
 
    cerebro.plot(style='candlestick')
 
 
if __name__ == '__main__':
    bt1()