获取股票数据主要是靠网页爬虫或者现成的库。
关于股票数据获取，python的接口有非常多，教程也有很多。
最后我选择了使用tushare和baostock。由于tushare升级之后有积分限制，很多数据是需要获取足够的积分才能得到，所以在学习量化交易的最开始使用了baostock获取数据。由于最后只需要获取沪深300和中证500的一些基础数据，所以也就联合baostock和tushare解决了这个问题(tushare获取沪深300和中证500成分股好像需要600积分)。实际上感觉tushare的接口比baostock更完善一些。
话不多说 直接上代码：
import tushare as tsimport osimport datetimeimport baostock as bsimport pandas as pdimport jsonimport loggerimport csvclass stockDump():def __init__(self, objLog=None, dataDir="stockData"):self.dictStock = {}self.dictMarket = {}self.objLog = objLogself.dataDir = dataDirself.maxLine = 0self.maxName = ""self.data_init()def data_init(self):if os.path.exists(self.dataDir) is False:os.makedirs(self.dataDir)def log(self, string):if self.objLog is None:print(string)else:self.objLog.addDrc(string)def login(self):bs.login()token = "your token"self.objTu = ts.pro_api(token)ts.set_token(token)def logout(self):bs.logout()def get_one_stock_kline(self, mark, startDate, endDate, dumpList=False):self.log("INFO: get info of '%s'"%(mark))rs = bs.query_history_k_data_plus(mark,"date,code,open,high,low,close,volume,amount,adjustflag,turn,pctChg,peTTM",start_date=startDate, end_date=endDate,frequency="d", adjustflag="2")data_list = []while (rs.error_code == '0') & rs.next():data_list.append(rs.get_row_data())result = pd.DataFrame(data_list, columns=rs.fields)self.maxLine = max(len(data_list), self.maxLine)if len(data_list) > self.maxLine:self.maxLine = len(data_list)self.maxName = markfName = mark.split(".")[-1]+".csv"fPath = os.path.join(self.dataDir, fName)result.to_csv(fPath, index=False)if dumpList is True:self.dictStock[mark] = "Add"def dump_stock_list(self):fPath = os.path.join(self.dataDir, "stocklist.js")fp = open(fPath, "w")json.dump(self.dictStock, fp)fp.close()def dump_industry(self):rs = bs.query_stock_industry()dictIndustry = {}while (rs.error_code == '0') & rs.next():rowData = rs.get_row_data()mark = rowData[1].split(".")[-1]dictIndustry[mark] = rowData[1:]fPath = os.path.join(self.dataDir, "industry.js")fp = open(fPath, "w")json.dump(dictIndustry, fp)fp.close()def get_his300_stocks(self, startDate, endDate, dumpList=True):rs = bs.query_hs300_stocks()#query_zz500_stockshs300_stocks = []while (rs.error_code == '0') & rs.next():hs300_stocks.append(rs.get_row_data())for item in hs300_stocks:mark = item[1]name = item[2]if dumpList is True:self.dictStock[mark] = nameself.get_one_stock_kline(mark, startDate, endDate)def get_zz500_stocks(self, startDate, endDate, dumpList=True):rs = bs.query_zz500_stocks()zz500_stocks = []while (rs.error_code == '0') & rs.next():zz500_stocks.append(rs.get_row_data())for item in zz500_stocks:mark = item[1]name = item[2]if dumpList is True:self.dictStock[mark] = nameself.get_one_stock_kline(mark, startDate, endDate)def align_csv(self, name=None):maxName = self.maxName if name is None else namemaxName = maxName.split(".")[-1]fName = maxName +".csv"objCsv = pd.read_csv(os.path.join(self.dataDir, fName))dateList = list(objCsv["date"])headerList = list(objCsv.columns.values)dateInd = headerList.index("date")codeInd = headerList.index("code")openInd = headerList.index("open")closeInd = headerList.index("close")highInd = headerList.index("high")lowInd = headerList.index("low")tmpLine = [0 for item in headerList]fPath = os.path.join(self.dataDir, "stocklist.js")fp = open(fPath, "r")dictStockList = json.load(fp)fp.close()for code in list(dictStockList.keys()):closePri = 0preFix = code.split(".")[0]code = code.split(".")[-1]pCode = "%s.%s"%(code, preFix.upper())if code == maxName:continuefPath = os.path.join(self.dataDir, "%s.csv"%(code))self.log("Align '%s'"%(fPath))objCsv = pd.read_csv(fPath)curDateList = list(objCsv["date"])#objNew = pd.DataFrame([[]], columns=headerList)tmpList = []addNum = 0for i, date in enumerate(dateList):if date in curDateList:lineList = list(objCsv.loc[i-addNum])closePri = lineList[closeInd]tmpList.append(lineList)continueaddNum += 1tmpLine[dateInd] = datetmpLine[codeInd] = pCodetmpLine[openInd] = closePritmpLine[closeInd] = closePritmpLine[highInd] = closePritmpLine[lowInd] = closePritmpList.append(tmpLine.copy())objNew = pd.DataFrame(tmpList, columns=headerList)objNew.to_csv(fPath, index=False)def get_his300_stocks_ts(self, startDate, endDate):rs = bs.query_hs300_stocks()hs300_stocks = []while (rs.error_code == '0') & rs.next():hs300_stocks.append(rs.get_row_data())for item in hs300_stocks:mark = item[1]name = item[2]self.dictStock[mark] = namemark = "%s.%s"%(mark.split(".")[-1], mark.split(".")[0].upper())startDate = startDate.replace("-", "")endDate = endDate.replace("-", "")self.get_one_stock_kline_ts(mark, startDate, endDate)def get_zz500_stocks_ts(self, startDate, endDate, dumpList=True):rs = bs.query_zz500_stocks()zz500_stocks = []while (rs.error_code == '0') & rs.next():zz500_stocks.append(rs.get_row_data())for item in zz500_stocks:mark = item[1]name = item[2]self.dictStock[mark] = namemark = "%s.%s"%(mark.split(".")[-1], mark.split(".")[0].upper())startDate = startDate.replace("-", "")endDate = endDate.replace("-", "")self.get_one_stock_kline_ts(mark, startDate, endDate)def get_one_stock_kline_ts(self, mark, startDate, endDate, dumpList=False):self.log("INFO: get info of '%s'"%(mark))markNum = mark.split(".")[0]dPath = os.path.join(self.dataDir, "%s.csv"%(markNum))objData = ts.pro_bar(ts_code=mark, adj='qfq', start_date=startDate, end_date=endDate)objData.rename(columns={"trade_date": "date"}, inplace=True)objData.rename(columns={"ts_code": "code"}, inplace=True)for i in range(0, len(objData)):date = datetime.datetime.strptime(objData.loc[i, "date"], "%Y%m%d")date = date.strftime("%Y-%m-%d")objData.loc[i, "date"] = dateindexList = list(reversed(objData["date"]))if self.maxLine < len(objData):self.maxLine = len(objData)self.maxName = markNumobjData = objData.set_index("date")objData = objData.reindex(index=indexList)objData.to_csv(dPath)if dumpList is True:self.dictStock[markNum] = "Add"if __name__ == "__main__":objDump = stockDump()objDump.login()endDate = str(datetime.datetime.today()).split()[0]objDump.dump_industry()objDump.get_his300_stocks_ts("-10-01", endDate)objDump.get_zz500_stocks_ts(startDate='-10-01', endDate=endDate)objDump.dump_stock_list()objDump.align_csv()objDump.logout()
需要注意几点：
1. token变量设置为用户自己的token，通过tushare官网获取
2. align_csv函数用于对齐股票数据，是由于个股存在停牌机制，停牌的时间段就没有数据，需要手动补充上去，否则不同股票的数据长度就会不一致，后续处理可能会有问题
3. 沪深300成分股和中证500成分股会不断更新
4. 注意是需要前复权的股价还是后复权
5. 获取的是日K数据
6. 历史前复权数据是会变化的，好像是遇到增发这种情况的时候，复权股票价格就会变化
