import numpy as np
import pandas as pd
import sklearn.model_selection 
security='000001.XSHE'
price=get_price(security, start_date=None, end_date='2019-8-1', 
                frequency='daily', fields=None, skip_paused=False, fq='pre', count=2000).close

#建立dataFrame数据模板，便于后面处理
data=pd.DataFrame()
data['close']=price
data['date']=price.index
data=data.reset_index(drop=True)

# 计算20日移动平均价格
# import time
# it=time.time()
# for i in data.index:
#     if i<19:
#         data['x']=None
#     else:
#         #非法赋值，会导致处理时间大幅延迟，和发生不可预知的风险，赋值用标准Pandas定位df.loc或者df.iloc
#         data['x'][i]=data['close'][i-19:i+1].mean()
# ct1=time.time()-it
# ct1

import time
it=time.time()
for i in data.index:
    if i<19:
        data['x']=None
    else:
        data.loc[i,'x']=data['close'][i-19:i+1].mean()
ct=time.time()-it

ct

# 计算路径效率
def alpha(data):
    for i in data.index:
        if i<9:
            data['a']=None
        else:
            b=np.abs(data.loc[i,'close']-data.loc[i-9,'close'])
            c=0
            for j in range(i-9,i):
                d=np.abs(data.loc[j+1,'close']-data.loc[j,'close'])
                d=d+c
                c=d
            data.loc[i,'a']=b/d
alpha(data)

#计算变换因子
def cur_b(data):
    if data.a.empty:
        data['b']=None
    else:
        data['b']=np.square(data.a*(2/3-2/14)+2/14)
    return data
data=cur_b(data)

# 以20日均线为基础的，自适应均线
def cur_data(data):
    for i in data.index:
        if i<19:
            data['y']=None
        elif i==19:
            data['y']=data['x']
        elif i>19:
            aa=data['b'][i]
            bb=data['x'][i]
            cc=data['y'][i-1]
            data.loc[i,'y']=aa*bb+(1-aa)*cc
cur_data(data)

# 以收盘价为基础的自适应均线
def cur_data1(data):
    for i in data.index:
        if i<19:
            data['y2']=None
        elif i==19:
            data['y2']=data['close']
        elif i>19:
            aa=data['b'][i]
            bb=data['close'][i]
            cc=data['y2'][i-1]
            data.loc[i,'y2']=aa*bb+(1-aa)*cc
cur_data1(data)

# 画图查看均线之间的关系
import matplotlib.pyplot as plt
plt.figure(figsize=(16,8))
plt.plot(data.date,data.x,label='ma')
plt.plot(data.date,data.y,label='zishiying')
plt.plot(data.date,data.y2,label='zishiying2')
#plt.plot(data.date,data.close,label='close')
plt.legend()
plt.show()

 从图形上看，绿线的变化速度要远快于黄色，可以利用绿线和黄线的变化来构建交易策略。

data