sma策略（用tushare包）

1. 简易版SMA策略一旦遇到股价震荡频率较高的股票，收益往往会出现问题，因此做个简单的策略升级。基于简易版，设置股价差20为阈值。当出现金叉且股价差大于20时再做多，当出现死叉且股价差小于20时再做空。分别计算大盘和策略的累积收益，并可视化为折线图，并与简易版作对比，看策略收益存在多大的区别
2. 使用动量交易策略分析，分别以自身（本身/2天/6天）的平均return作为position，生成strategy再分别做可视化折线图
   简易版为下图 （回答附上code 感谢）

在原来计算出移动平均数的基础上，增加一些判断条件进行计算并画出对比图。参考如下代码：

#无策略
df['date_rev'] = [0]+[(df.loc[i,'close']-df.loc[i-1,'close'])
                      for i in range(1, len(df))]
df['ac_rev'] = df['date_rev'].cumsum()
#第一策略
a=[0]
for i in range(1, len(df)):
    if df.loc[i-1, 'm10'] < df.loc[i-1, 'm60'] and df.loc[i, 'm10'] > df.loc[i, 'm60']:
        n = i
    if df.loc[i-1, 'm10'] > df.loc[i-1, 'm60'] and df.loc[i, 'm10'] < df.loc[i, 'm60']:
        a.append(df.loc[i, 'close']-df.loc[n, 'close'])
    else:
        a.append(0)
df['st1_rev']=a
df['ac_st1_rev']=df['st1_rev'].cumsum()
#第二策略
x = [0]
for i in range(1,len(df)):
    if df.loc[i-1, 'm10'] < df.loc[i-1, 'm60'] and df.loc[i, 'm10'] > df.loc[i, 'm60'] and abs(df.loc[i,'close']-df.loc[i-1,'close']) >= 20:
        n = i
    if df.loc[i-1, 'm10'] > df.loc[i-1, 'm60'] and df.loc[i, 'm10'] < df.loc[i, 'm60'] and abs(df.loc[i, 'close']-df.loc[i-1, 'close']) >= 20:
        x.append(df.loc[i, 'close']-df.loc[n, 'close'])
    else:
        x.append(0)
#绘制图
df['st2_rev'] = x
df['ac_st2_rev'] = df['st2_rev'].cumsum()
df.to_csv('hs300_2.csv')
plt.figure(figsize=(12,8))
plt.subplot(1,2,1)
plt.title('strategy1')
plt.plot(df['date'], df['ac_rev'],df['ac_st1_rev'])
plt.legend(['ac_rev','ac_st1_rev'])
plt.subplot(1,2,2)
plt.title('strategy2')
plt.plot(df['date'], df['ac_rev'],df['ac_st2_rev'])
plt.legend(['ac_rev', 'ac_st2_rev'])
plt.show()

均价计算及折线作图仅可参考，至于策略开发，需根据每个交叉点前后两个均价的比较写个判断，采取不同的sell和buy不同交易方法。

import tushare as ts
import pandas as pd
df = ts.get_k_data('hs300', start='2014-06-30', end='2021-06-30')
#df=pd.read_csv('hs300.csv')
pd.set_option('max_rows',None)
a10=[]
a60=[]
for i in range(len(df)):
    if i<=9:
        a10.append(df['close'][:i+1].mean())
    else:
        a10.append(df['close'][i-10:i].mean())
    if i<=59:
        a60.append(df['close'][:i+1].mean())
    else:
        a60.append(df['close'][i-60:i].mean())
df['a10']=a10
df['a60']=a60
print(df.head(20))
m = 1000
rev0 = m * (df['close'].tolist()[-1]-df['close'].tolist()[0]) / \
    df['close'].tolist()[0]-m#不采取交易策略下的收益
import matplotlib.pyplot as plt
plt.plot(df['date'],df['a10'],label='ma10')
plt.plot(df['date'], df['a60'], label='ma60')
plt.legend()
plt.show()