macd1.py

import baostock as bs
import pandas as pd
import talib as ta
import matplotlib.pyplot as plt

def computeMACD(code,startdate,enddate):
    
    login_result = bs.login(user_id='anonymous', password='123456')
    print(login_result)
    
    ###获取股票日K线数据###
    rs = bs.query_history_k_data(code,
        "date,code,close,tradeStatus",
        start_date=startdate, end_date=enddate, 
        frequency="d", adjustflag="3")
    #### 打印结果集 ####
    result_list = []
    while (rs.error_code == '0') & rs.next():
        # 获取一条记录，将记录合并在一起
        result_list.append(rs.get_row_data())
    df = pd.DataFrame(result_list, columns=rs.fields)
    #剔除停盘数据
    df2 = df[df['tradeStatus']=='1']
    #获取dif,dea,hist，它们的数据类似是tuple，且跟df2的date日期一一对应
    #记住了dif,dea,hist前33个为Nan，所以推荐用于计算的数据量一般为你所求日期之间数据量的3倍
    #这里计算的hist就是dif-dea,而很多证券商计算的MACD=hist*2=(dif-dea)*2
    dif, dea, hist = ta.MACD(df2['close'].astype(float).values, fastperiod=12, slowperiod=26, signalperiod=9)
    df3 = pd.DataFrame({'dif':dif[33:],'dea':dea[33:],'hist':hist[33:]},
                       index=df2['date'][33:],columns=['dif','dea','hist'])
    df3.plot(title='MACD')
    plt.show()
    #寻找MACD金叉和死叉
    datenumber = int(df3.shape[0])
    for i in range(datenumber-1):
        if ((df3.iloc[i,0]<=df3.iloc[i,1]) & (df3.iloc[i+1,0]>=df3.iloc[i+1,1])):
            print("MACD金叉的日期："+df3.index[i+1])
        if ((df3.iloc[i,0]>=df3.iloc[i,1]) & (df3.iloc[i+1,0]<=df3.iloc[i+1,1])):
            print("MACD死叉的日期："+df3.index[i+1])

    bs.logout()
    return(dif,dea,hist)

 
if __name__ == '__main__':
    code = 'sh.600004'
    startdate = '2017-03-01'
    enddate = '2020-04-13'
    (dif,dea,hist) = computeMACD(code,startdate,enddate)