Github

https://github.com/micovey/Quantify

1. 代碼目的
設(shè)定一定的時(shí)間段，滾動(dòng)下載數(shù)據(jù)。
例如，當(dāng)前時(shí)間為2019-11-19，當(dāng)日未收盤，若設(shè)置數(shù)據(jù)期間為180天，則保存2019-5-23至2019-11-18的數(shù)據(jù)
明日為2019-11-20，當(dāng)日未收盤，則自動(dòng)刪除2019-5-23數(shù)據(jù)，加入2019-11-19的數(shù)據(jù)。
便于之后的量化分析
滾動(dòng)的目的：避免數(shù)據(jù)文件過大
2. 前期準(zhǔn)備
首先安裝python （或者anaconda），但要注意最好不要同時(shí)安裝python和anacoda，因?yàn)閍nacoda里面包含純python，容易起沖突。
anacoda網(wǎng)址：https://www.anaconda.com/
若選擇純python， 則需要

pip install datetime
pip install pandas
pip install numpy
pip install baostock

若選擇anacoda

需要設(shè)置環(huán)境變量

111111.png

然后只需要

pip install baostock

安裝pycharm或者anacoda自帶的spyder

3. 為什么選擇baostock

Tushare、Baostock、joinquant等均可獲得股票日數(shù)據(jù)
Wind、CSMAR也可以獲得（花錢）
但是Tusharepro需要積分、joinquant只有一年試用期
因此本文選擇Baostock，無需注冊(cè)。
http://baostock.com/

4. 代碼

import baostock as bs
import pandas as pd
import numpy as np
import datetime
#### 登陸系統(tǒng) ####
lg = bs.login()
########定義變量#######
now_time=datetime.datetime.now().strftime('%Y-%m-%d')
now_time=datetime.datetime.strptime(now_time,'%Y-%m-%d')
year=int(now_time.year)
print('請(qǐng)輸入地址')
global file_place
file_place='D:\\'
fl = 'Quantify\\idcode.csv'
filee=file_place+fl
fff=file_place+'\\Quantify\\daily\\daily'+str(year)+'.csv'
# D:\\Quantify\\idcode.csv
idcode = pd.read_csv(filee,header=None)
if len(idcode) > 10:
    print('讀取成功')
else:
    print('讀取失敗')

idcode = np.array(idcode)
global datalength
datalength=180
#######定義函數(shù)########

##滾動(dòng)刪除數(shù)據(jù)

def drop_date_row(drop_data):
    drop_data['date'] = [datetime.datetime.strptime(x, '%Y/%m/%d') for x in drop_data['date']]
    now_time=datetime.datetime.now().strftime('%Y-%m-%d')
    now_time=datetime.datetime.strptime(now_time,'%Y-%m-%d')
    min_date = drop_data['date'].min()
    drop_date = now_time + datetime.timedelta(days=-datalength)
    if min_date <drop_date:
        drop_data = drop_data.drop(drop_data[drop_data.date < drop_date].index)
        return(drop_data)
    else:

        drop_data=pd.DataFrame()
        return(drop_data )

##滾動(dòng)獲得數(shù)據(jù)

def get_daily_data(id,start_date_find,end_date_find):
    rs = bs.query_history_k_data_plus(id,        "date,code,open,high,low,close,preclose,volume,amount,adjustflag,turn,tradestatus,pctChg,isST,peTTM,pbMRQ,psTTM,pcfNcfTTM",
        start_date=str(start_date_find), end_date=str(end_date_find),
        frequency="d", adjustflag="3")
    data_list = []
    while (rs.error_code == '0') & rs.next():
        data_list.append(rs.get_row_data())
    result = pd.DataFrame(data_list, columns=rs.fields,index=range(0,len(data_list)))
    result["turn"] = [0 if x == "" else float(x) for x in result["turn"]]
    result["volume"] = [0 if x == "" else float(x) for x in result["volume"]]
    result["close"] = [0 if x == "" else float(x) for x in result["close"]]
    marketvalue=result["volume"] / result["turn"]*result["close"]*100
  #  marketvalue=result.apply(lambda x: x["volume"] / x["turn"]*x["close"]*100, axis=1)
    result["marketvalue"]=marketvalue
#  print( result)
    return(result)
##開始，結(jié)束下載日期

def check_download(check_date):
    if len(check_date)==0:
        now_time = datetime.datetime.now().strftime('%Y-%m-%d')
        now_time = datetime.datetime.strptime(now_time, '%Y-%m-%d')
        start_date = now_time + datetime.timedelta(days=-datalength)
        start_date = str(datetime.datetime.strftime(start_date, '%Y-%m-%d'))
    else:
        start_date = check_date['date'].max()
        start_date += datetime.timedelta(days=+1)
        start_date = str(datetime.datetime.strftime(start_date, '%Y-%m-%d'))
    end_date=str(datetime.datetime.now().strftime('%Y-%m-%d'))
    return start_date,end_date

##刪除重復(fù)項(xiàng)

def drop_duplicates(drop_data):
    n1=len(drop_data)
    drop_data['date'] = [datetime.datetime.strptime(x, '%Y/%m/%d') for x in drop_data['date']]
    drop_data.drop_duplicates(subset=['code','date'],keep='first',inplace=True)
    n2=len(drop_data)
    if n1==n2:
        drop_data = pd.DataFrame()
        return (drop_data)
    else:
        return (drop_data)

#####開始計(jì)算###########

######滾動(dòng)刪除數(shù)據(jù)

drop_data = pd.read_csv(fff)
drop_data = drop_date_row(drop_data)
if len(drop_data)<2:
    print("不需要?jiǎng)h除數(shù)據(jù)")
else:
    print("已刪除數(shù)據(jù)")
    drop_data['date'] = [datetime.datetime.strftime(x, '%Y/%m/%d') for x in drop_data['date']]
    drop_data.to_csv(fff, header=True, index=False)

##查看是否需要更新數(shù)據(jù)
check_date = pd.read_csv(fff)
check_date['date']=[datetime.datetime.strptime(x,'%Y/%m/%d') for x in check_date['date']]
start_date,end_date=check_download(check_date)
##歷史行情數(shù)據(jù)下載

if datetime.datetime.strptime(start_date,'%Y-%m-%d')<=datetime.datetime.strptime(end_date,'%Y-%m-%d'):
    if datetime.datetime.strptime(end_date,'%Y-%m-%d')==datetime.datetime.strptime(start_date,'%Y-%m-%d') and int(datetime.datetime.now().hour)<18:
        print("不需要更新數(shù)據(jù)，當(dāng)日沒有結(jié)束")
    else:
        for id in idcode:
            id=str(id)
            id=id[2:11]#提取9位代碼
            print('=====' + id + '=====')
            result1=get_daily_data(id,start_date,end_date)
            if len(result1) > 0:
                result1['date'] = [datetime.datetime.strptime(x,'%Y-%m-%d') for x in result1['date']]
                result1['quarter']=[int((x.month-1)/3)+1 for x in result1['date']]
                result1['year']=[int(x.year) for x in result1['date']]
                result1['date'] = [datetime.datetime.strftime(x, '%Y/%m/%d') for x in result1['date']]
                result1.to_csv(fff, mode='a', header=False)
            else:
                continue
else:
    print("不需要更新數(shù)據(jù)，沒有開始新交易")

##刪除重復(fù)項(xiàng)

drop_data = pd.read_csv(fff)
drop_data = drop_duplicates(drop_data)
if len(drop_data)<2:
    print("沒有重復(fù)項(xiàng)")
else:
    print("已刪除重復(fù)項(xiàng)")
    drop_data['date'] = [datetime.datetime.strftime(x, '%Y/%m/%d') for x in drop_data['date']]
    drop_data.to_csv(fff, header=True, index=False)
##
bs.logout()