def get_daily_data(self, code, expire=60 * 6):
"""
获取一支股票所有历史数据保存到本地
"""
UT.check_dir(CT.DAILY_DIR)
file_path = CT.DAILY_DIR + code
expired = UT.check_file_expired(file_path, expire)
if expired or not os.path.exists(file_path):
symbol = self._code_to_symbol(code)
start_date = CT.START
end_date = date_time.get_today_str()
adjust = 'qfq'
if is_index(code):
d = ak.stock_zh_index_daily(symbol)
else:
d = ak.stock_zh_a_daily(symbol, start_date, end_date, adjust)
if d is None:
return d
d.to_csv(file_path, sep='\t')
if not os.path.exists(file_path):
return None
#d = pd.read_csv(file_path, sep='\t', index_col=0)
d = pd.read_csv(file_path,
sep='\t',
skiprows=0,
parse_dates=True,
header=0,
index_col=0)
return d
def get_quotation(self, stock_code):
''' 从网络上获取行情数据 '''
stock_symbol = 'sh{0}'.format(stock_code)
stock_df = ak.stock_zh_a_daily(symbol=stock_symbol, factor="")
print('df:{0}; {1}'.format(type(stock_df), stock_df))
stock_df.to_csv(path_or_buf='./data/tp/{0}.csv'.format(stock_symbol),
sep=',')
def get_data_from_internet(self, code):
'''获取数据并保存到csv'''
# 获取原始数据
original_data = ak.stock_zh_a_daily(symbol=code, adjust="qfq")
# 取过去30天数据
df = original_data.reset_index().iloc[:, :6]
# 去除空值且从零开始编号索引
df = df.dropna(how='any').reset_index(drop=True)
# 按日期排序
df = df.sort_values(by='date', ascending=True)
# 均线数据
df['10'] = df.close.rolling(10).mean()
df['60'] = df.close.rolling(60).mean()
df['250'] = df.close.rolling(250).mean()
# 写入csv
body_json = ReadConfig().read_json("code_list.json")
first_dict = body_json["first_team"]
second_dict = body_json["second_team"]
if code in first_dict:
file_name = self.csv_path + first_dict.get(code) + ".csv"
elif code in second_dict:
file_name = self.csv_path + second_dict.get(code) + ".csv"
else:
file_name = self.csv_path + code + ".csv"
df.to_csv(file_name)
return df
def get_historical_news(self, start_date=None, end_date=None, freq="day"):
stock_symbol_list = self.col_basic_info.distinct("symbol")
if len(stock_symbol_list) == 0:
stock_symbol_list = self.get_stock_code_info()
if freq == "day":
if os.path.exists(config.STOCK_DAILY_EXCEPTION_TXT_FILE_PATH):
with open(config.STOCK_DAILY_EXCEPTION_TXT_FILE_PATH,
"r") as file:
start_stock_code = file.read()
logging.info(
"read {} to get start code number is {} ... ".format(
config.STOCK_DAILY_EXCEPTION_TXT_FILE_PATH,
start_stock_code))
else:
start_stock_code = 0
for symbol in stock_symbol_list:
if int(symbol[2:]) >= int(start_stock_code):
try:
# TODO
if start_date is None:
# 如果该symbol有历史数据，如果有则从API获取从数据库中最近的时间开始直到现在的所有价格数据
# 如果该symbol无历史数据，则从API获取从2015年1月1日开始直到现在的所有价格数据
start_date = "20150101"
if end_date is None:
pass
stock_zh_a_daily_hfq_df = ak.stock_zh_a_daily(
symbol=symbol,
start_date=start_date,
end_date=end_date,
adjust="hfq")
stock_zh_a_daily_hfq_df.insert(
0, 'date', stock_zh_a_daily_hfq_df.index.tolist())
stock_zh_a_daily_hfq_df.index = range(
len(stock_zh_a_daily_hfq_df))
_col = self.db_obj.get_collection(
self.database_name, symbol)
for _id in range(stock_zh_a_daily_hfq_df.shape[0]):
_col.insert_one(
stock_zh_a_daily_hfq_df.iloc[_id].to_dict())
logging.info("{} finished saving ... ".format(symbol))
except Exception:
with open(config.STOCK_DAILY_EXCEPTION_TXT_FILE_PATH,
"w") as file:
file.write(symbol[2:])
elif freq == "week":
pass
elif freq == "month":
pass
elif freq == "5mins":
pass
elif freq == "15mins":
pass
elif freq == "30mins":
pass
elif freq == "60mins":
pass
def k_line(symble, title, last=-30, adjust="qfq", dprint=False):
'''
过去n天个股历史行情均线数据（k线）
param:
symble: str,股票编号"sh601727"
title: str,作图标题
last: int,过去多少天的数据,默认过去30天，例"-30"
adjust: str，默认"qfq"前复权,返回复权后数据
dprint: bool,默认false,是否打印原始数据
startday: int,行情查询起始日期（懒得转化，没写功能）
endday: int,行情查询起始日期（懒得转化，没写功能）
'''
pingan = ak.stock_zh_a_daily(symble, adjust)
df3 = pingan.reset_index().iloc[last:, :6] #取过去30天数据
df3 = df3.dropna(how='any').reset_index(drop=True) #去除空值且从零开始编号索引
df3 = df3.sort_values(by='date', ascending=True)
print(df3.info())
# 均线数据
df3['5'] = df3.close.rolling(5).mean()
df3['10'] = df3.close.rolling(10).mean()
df3.tail()
#画图
fig, ax = plt.subplots(1, 1, figsize=(8, 3), dpi=200)
candlestick2_ohlc(ax,
opens=df3['open'].values,
highs=df3['high'].values,
lows=df3['low'].values,
closes=df3['close'].values,
width=0.5,
colorup="r",
colordown="g")
# 显示最高点和最低点
ax.text(df3.high.idxmax(), df3.high.max(), s=df3.high.max(), fontsize=8)
ax.text(df3.high.idxmin(),
df3.high.min() - 2,
s=df3.high.min(),
fontsize=8)
ax.set_facecolor("white")
ax.set_title(title)
# 画均线
plt.plot(df3['5'].values, alpha=0.5, label='MA5')
plt.plot(df3['10'].values, alpha=0.5, label='MA10')
ax.legend(facecolor='white', edgecolor='white', fontsize=6)
# 修改x轴坐标
plt.xticks(ticks=np.arange(0, len(df3)),
labels=df3.date.dt.strftime('%Y-%m-%d').to_numpy())
plt.xticks(rotation=90, size=4)
# 修改y轴坐标
ax.yaxis.set_major_formatter(FormatStrFormatter('%.2f'))
plt.show()
def get_stockQuotes(self, stockcode, stardate, enddate):
if stockcode[0:2] == '60' or stockcode[0:2] == '68':
stockcode = 'sh' + stockcode
else:
stockcode = 'sz' + stockcode
response = ak.stock_zh_a_daily(symbol=stockcode,
start_date=stardate,
end_date=enddate,
adjust="qfq")
return response
def startup(self, code='sh601318', start_cash=100000.0):
print('回测系统示例程序 v0.0.2')
cerebro = bt.Cerebro()
cerebro.addstrategy(TestStrategy)
stock_zh_a_daily_df = ak.stock_zh_a_daily(
symbol=code, adjust="hfq") # 通过 AkShare 获取需要的数据
data = bt.feeds.PandasData(dataname=stock_zh_a_daily_df) # 规范化数据格式
cerebro.adddata(data) # 将数据加载至回测系统
cerebro.broker.setcash(start_cash) # 设置初始资金
cerebro.broker.set_commission_obj(AshareCommInfo())
print('期初净值: {0:.2f}'.format(cerebro.broker.getvalue()))
cerebro.run()
print('期末净值: {0:.2f}'.format(cerebro.broker.getvalue()))
def main(code="sh601318", start_cash=1000000, stake=100, commission_fee=0.001):
cerebro = bt.Cerebro() # 创建主控制器
cerebro.optstrategy(MyStrategy, maperiod=range(3, 31)) # 导入策略参数寻优
stock_zh_a_daily_df = ak.stock_zh_a_daily(
symbol=code, adjust="hfq") # 通过 AkShare 获取需要的数据
data = bt.feeds.PandasData(dataname=stock_zh_a_daily_df) # 规范化数据格式
cerebro.adddata(data) # 将数据加载至回测系统
cerebro.broker.setcash(start_cash) # broker设置资金
cerebro.broker.setcommission(commission=commission_fee) # broker手续费
cerebro.addsizer(bt.sizers.FixedSize, stake=stake) # 设置买入数量
print("期初总资金: %.2f" % cerebro.broker.getvalue())
cerebro.run(maxcpus=1) # 用单核 CPU 做优化
print("期末总资金: %.2f" % cerebro.broker.getvalue())
def __init__(self, symbol):
self.name = 'fas.bktr.MarketDataSource'
self.event_tick = None
self.symbol = symbol # 'sh600582'
self.market_data = MarketData()
datas = ak.stock_zh_a_daily(symbol=self.symbol, adjust='hfq')
self.open_pds = datas['open']
self.high_pds = datas['high']
self.low_pds = datas['low']
self.close_pds = datas['close']
self.volume_pds = datas['volume']
self.outstanding_share_pds = datas['outstanding_share']
self.turnover_pds = datas['turnover']
def download(self):
for type in ['trade','favourite']:
tickers = pd.read_csv(self.BASIC + type +'.csv', encoding='gb18030')['tag'].values.tolist()
alltickers = tickers.copy()
removed_tickers = []
# tickers = list(set(tickers).difference(set([i[:-4] for i in os.listdir(self.AKSHARE + type + '/') ])))
for ticker in tickers:
if ticker[:2]=='E_':
try:
data = ak.fund_etf_hist_sina(symbol=ticker[2:])
data = data.drop_duplicates()
data.to_csv(self.AKSHARE + type + '/' + ticker + '.csv', encoding='gb18030', index=False)
except :
print('Can not download ETF:',ticker)
removed_tickers.append(ticker)
elif ticker[:2]=='S_':
try:
data = ak.stock_zh_a_daily(symbol=ticker[2:], adjust="qfq")
data.drop(['outstanding_share','turnover'],axis=1,inplace=True)
data = data.drop_duplicates()
data.to_csv(self.AKSHARE + type + '/' + ticker + '.csv', encoding='gb18030', index=True)
except :
print('Can not download stock:',ticker)
removed_tickers.append(ticker)
elif ticker[:2]=='I_':
try:
data = ak.stock_zh_index_daily(symbol=ticker[2:])
data = data.drop_duplicates()
data.to_csv(self.AKSHARE + type + '/' + ticker + '.csv', encoding='gb18030', index=True)
except :
print('Can not download index:',ticker)
removed_tickers.append(ticker)
for ticker in removed_tickers: # Remove empty record
try:
os.remove(raw_path + ticker + '.csv')
except:
pass
if type == 'favourite':
data = pd.DataFrame(list(set(alltickers).difference(set(removed_tickers))),columns =['tag'])
data.sort_values(by=['tag'],ascending=[True],inplace=True)
data.to_csv(self.BASIC + 'favourite.csv', encoding='gb18030',index=False)
elif type == 'archive':
data = pd.DataFrame(list(set(alltickers).difference(set(removed_tickers))),columns =['tag'])
data.sort_values(by=['tag'],ascending=[True],inplace=True)
data.to_csv(self.BASIC + 'archive.csv', encoding='gb18030',index=False)
return
def insertTodayValue(data,table):
today = datetime.datetime.now()
configger.init()
engine=configger.engine
i=0
symbols = tqdm(data['symbol'])
now=datetime.datetime.now()
sql='select code,max(date) as date,max(updateTime)as updateTime from {} GROUP BY code'.format(table)
# print(sql)
try:
timeData=pd.read_sql(con=engine,sql=sql,index_col='code')
except:
timeData=pd.DataFrame()
# print(timeData)
for code in symbols:
code=util.removedotBysymbol(code)
i=i+1
# code=util.getdotCodeBysymbol(symbol)
try:
if(not timeData.empty and code in timeData.index and (now-timeData.loc[code,'updateTime']).seconds<3600*24):
continue
except:
pass
try:
start_date= pd.to_datetime(timeData.loc[code,'date']) + datetime.timedelta(days=1)
start_date=start_date.strftime('%Y%m%d')
# start_date=start_date.date()
except:
start_date='19900101'
# traceback.print_exc()
end_date=today.strftime('%Y-%m-%d')
# print(start_date,end_date)
if(start_date>end_date):
# print(code,start_date,'已存在')
continue
# print('star_Date')
# print(start_date)
try:
result=ak.stock_zh_a_daily(symbol=code, start_date=start_date, end_date=end_date, adjust="qfq")
result.reset_index(inplace=True)
result.drop(columns='index',inplace=True)
result['updateTime']=now
result['code']=code
print(result)
result.to_sql(table,con=engine,if_exists='append',index=False)
symbols.set_description("查询代码为：{},数据条数为{}".format(code,len(result.index)))
except:
traceback.print_exc()
def query_stock_codes_save():
stock_infos = DBUtils.execute("select * from stock_2020_1218")
for data in stock_infos:
print(data['symbol'])
stock_zh_a_daily_hfq_df = ak.stock_zh_a_daily(symbol=data['symbol'],
adjust="qfq")
for index, stock_price in stock_zh_a_daily_hfq_df.iterrows():
if stock_price["open"] == 'nan' or stock_price["volume"] == 'nan':
continue
save_stock_price_info({
"date":
index,
**dict(stock_price), "stock_code":
data['symbol']
})
def addData(self, code, name):
tmp_stock = ""
if int(code) > 600000:
tmp_stock = "sh" + code
else:
tmp_stock = "sz" + code
stodk_data_daily = ak.stock_zh_a_daily(symbol=tmp_stock, adjust="hfq")
print(stodk_data_daily)
mydb = mysql.connector.connect(host="localhost",
user="******",
password="******",
database="mydatabase",
charset='utf8')
mycursor = mydb.cursor()
for i in range(0, len(stodk_data_daily)):
# SQL 插入语句
data = []
date = [t.strftime("%Y-%m-%d") for t in stodk_data_daily.index][i]
open = stodk_data_daily.iat[i, 0]
high = stodk_data_daily.iat[i, 1]
low = stodk_data_daily.iat[i, 2]
close = stodk_data_daily.iat[i, 3]
volume = stodk_data_daily.iat[i, 4]
data.append(date)
data.append(open)
data.append(high)
data.append(low)
data.append(close)
data.append(volume)
self.stock_datas.append(data)
# sql = "select * from stock_detail \
# where code = '%s' and data = '%s'" % (code,date)
sql = "insert into stock_detail(code, name, date, open, high, low, close, volume) \
values('%s','%s','%s','%f','%f','%f','%f','%f')" % \
(code, name, date, open, high,low,close,volume)
try:
# 执行sql语句
mycursor.execute(sql)
# 提交到数据库执行
mydb.commit()
except:
# Rollback in case there is any error
mydb.rollback()
mydb.close()
def query_bond_stock_codes_save():
bond_infos = DBUtils.execute(
"select 正股代码, 申购日期, 交易场所 from kezhuanzhai "
"where 申购日期>'2015-01-01' and 申购日期< '2020-11-01' order by 正股代码")
for data in bond_infos:
stock_code = str(data["交易场所"][-2:]).lower() + data['正股代码']
print(data['正股代码'])
stock_zh_a_daily_hfq_df = ak.stock_zh_a_daily(symbol=stock_code,
adjust="qfq")
for index, stock_price in stock_zh_a_daily_hfq_df.iterrows():
if stock_price["open"] == 'nan' or stock_price["volume"] == 'nan':
continue
save_stock_price_info({
"date":
index,
**dict(stock_price), "stock_code":
data['正股代码']
})
def price_trend(
stock_code,
start="20160101",
end=datetime.datetime.now().strftime("%Y%m%d")
) -> pd.DataFrame:
data_file = os.path.join(price_cache_dir(), stock_code + '.csv')
try:
price_trend_df = pd.read_csv(data_file, index_col=0)
price_trend_df.index = pd.to_datetime(price_trend_df.index)
return price_trend_df
except FileNotFoundError:
stock_a_indicator_df = ak.stock_a_lg_indicator(stock=stock_code[-6:])
stock_a_indicator_df['trade_date'] = pd.to_datetime(stock_a_indicator_df['trade_date'])
stock_a_indicator_df['date'] = stock_a_indicator_df['trade_date']
stock_zh_a_daily_qfq_df = ak.stock_zh_a_daily(symbol=stock_code, start_date=start, end_date=end, adjust="qfq")
price_trend_df = pd.merge(stock_zh_a_daily_qfq_df, stock_a_indicator_df, how='left', on=['date'])
price_trend_df.set_index(['date'], drop=True, inplace=True)
price_trend_df.to_csv(data_file)
return price_trend_df
def get_data_from_internet(self, code, name):
'''获取数据并保存到csv'''
while len(code) < 6:
code = '0' + code
if code.startswith('6'):
code = "sh" + code
if code.startswith('3') or code.startswith('0'):
code = "sz" + code
print(code)
# 获取原始数据
original_data = ak.stock_zh_a_daily(symbol=code, adjust="qfq")
# 取过去600天数据
try:
df = original_data.reset_index().iloc[-600:, :6]
except:
df = original_data.reset_index().iloc[:, :6]
# 去除空值且从零开始编号索引
df = df.dropna(how='any').reset_index(drop=True)
#print(df)
# 按日期排序
df = df.sort_values(by='date', ascending=True)
# 均线数据
# df['10'] = df.close.rolling(10).mean()
# df['60'] = df.close.rolling(60).mean()
# df['250'] = df.close.rolling(250).mean()
df['120'] = df.close.rolling(120).mean()
df['180'] = df.close.rolling(180).mean()
df['240'] = df.close.rolling(240).mean()
# 写入csv
file_name = self.data_path + name + ".csv"
df.to_csv(file_name)
print("<" + name + ">数据完成")
return df
def create_df(self):
"""创建df模板"""
if (len(self.symbols) == 0):
log.error("股票列表为空，无法创建dataframe")
return
try:
time_start = time.time()
data = ak.stock_zh_a_daily(symbol=self.symbols[0], adjust="hfq")
time_end = time.time()
log.debug("查询一次数据时间(只测了一次，仅做参考):{0}s".format(time_end -
time_start))
except Exception as e:
log.error("取不到此股票的历史数据,symbol:{0}".format(self.symbols[0]))
log.error(e)
return
log.debug("行索引个数:{0}".format(len(data.index)))
self.open_df = pd.DataFrame(index=data.index)
self.high_df = pd.DataFrame(index=data.index)
self.low_df = pd.DataFrame(index=data.index)
self.close_df = pd.DataFrame(index=data.index)
self.turnover_df = pd.DataFrame(index=data.index)
self.outstanding_df = pd.DataFrame(index=data.index)
log.debug("创建dataframe成功")
def extract_stock_records(code='sh600000', start=None, end =None,adjust = 'qfq', identifier = 'a'):
#qfq:前复权, hfq:后复权
if identifier == 'a':
result = ak.stock_zh_a_daily(symbol=code, adjust=adjust)
elif identifier == 'hk':
result = ak.stock_hk_daily(symbol=code, adjust=adjust)
elif identifier == 'us':
result = ak.stock_us_daily(symbol=code, adjust=adjust)
result.index = result.index.tz_localize(None)
result['date'] = result.index
result = result[['date','open','close','high','low','volume']]
result['total'] = 'NaN'
result['amp'] = 'NaN'
result.columns = ['date','open_price','close_price','high_price','low_price','trancaction','total','amp']
result['rate'] = pd.Series([0]).append(pd.Series((result['close_price'].iloc[1:].values-result['close_price'].iloc[0:-1].values)/result['close_price'].iloc[0:-1].values)).values*100
if start==None and end==None:
pass
else:
result = result[(result['date']<=pd.Timestamp(end)) & (result['date']>=pd.Timestamp(start))]
result['date'] = result['date'].apply(lambda x: (x.date()-datetime.date(1, 1, 1)).days)
result.index = result.index.rename('')
result = result.reset_index()
return result[['date','open_price','close_price','low_price','high_price','trancaction','total','rate','amp']]
def get_history_data(self, adjust=""):
"""获取一种类型的历史数据"""
self.adjust_type = adjust
self.get_symbols()
self.create_df()
total = len(self.symbols)
for index, symbol in enumerate(self.symbols):
try:
data = ak.stock_zh_a_daily(
symbol=symbol, adjust=self.adjust_type) #adjust="hfq"
count = index + 1
if (count % 1 == 0):
log.debug("当前获取进度{:.2%}:共{},第{}".format(
count / total, total, count))
self.parse_data(data, symbol)
if (count == 10): break
except Exception as e:
if (e.args[0] == "No value to decode"):
log.debug("已经退市:{0}".format(symbol))
else:
log.error("得到历史数据出错:{0}, {1}".format(symbol, e))
self.save_csv()
self.save_pickle()
def get_kline(code, start_date: datetime.date):
"""
返回json数据，最新的日期在前面
"""
result = []
df = ak.stock_zh_a_daily(symbol=get_type(code) + code, adjust="hfq")
# 腾讯，主要是使用amount字段
dftx = ak.stock_zh_index_daily_tx(symbol=get_type(code) + code)
if df is None or df.empty:
return result
if dftx is None or dftx.empty:
return result
for row in df.itertuples():
w = {}
w['datetime'] = getattr(row, 'Index')
if w['datetime'] < start_date:
continue
w['open'] = getattr(row, 'open')
w['high'] = getattr(row, 'high')
w['low'] = getattr(row, 'low')
w['close'] = getattr(row, 'close')
w['volume'] = getattr(row, 'volume')
result.append(w)
return result
elif self.crossover < 0:
self.log('创建卖单')
self.sell(size=100)
##########################
# 主程序开始
#########################
# 创建大脑引擎对象
cerebro = bt.Cerebro()
# start_date = datetime(2018, 1, 1) # 回测开始时间
# end_date = datetime(2020, 1, 1) # 回测结束时间
# 利用 AkShare 获取后复权数据
stock_hfq_df = ak.stock_zh_a_daily(symbol="sh600000", adjust="hfq",start_date='20180101',end_date='20200101')
# data = bt.feeds.PandasDirectData(dataname=stock_hfq_df, fromdate=start_date, todate=end_date) # 加载数据
print(stock_hfq_df)
data = bt.feeds.PandasDirectData(dataname=stock_hfq_df) # 加载数据
cerebro.adddata(data) # 将数据传入回测系统
cerebro.addstrategy(SmaCross) # 将交易策略加载到回测系统中
cerebro.broker.setcash(1000000.0) # 设置初始资金
cerebro.run() # 运行
print('最终市值: %.2f' % cerebro.broker.getvalue())
def stock_zh_a_daily_all(self, symbol):
ret = {}
temp_df = ak.stock_zh_a_daily(symbol=symbol, adjust="")
ret["defaut"] = temp_df
# "hfq":
temp_df_hfq = temp_df.copy()
res = requests.get(zh_sina_a_stock_hfq_url.format(symbol))
hfq_factor_df = pd.DataFrame(
eval(res.text.split("=")[1].split("\n")[0])['data'])
hfq_factor_df.columns = ["date", "hfq_factor"]
hfq_factor_df.index = pd.to_datetime(hfq_factor_df.date)
del hfq_factor_df["date"]
temp_df_hfq = pd.merge(temp_df_hfq,
hfq_factor_df,
left_index=True,
right_index=True,
how="left")
temp_df_hfq.fillna(method="ffill", inplace=True)
temp_df_hfq = temp_df_hfq.astype(float)
temp_df_hfq["open"] = temp_df_hfq["open"] * temp_df_hfq["hfq_factor"]
temp_df_hfq["high"] = temp_df_hfq["high"] * temp_df_hfq["hfq_factor"]
temp_df_hfq["close"] = temp_df_hfq["close"] * temp_df_hfq["hfq_factor"]
temp_df_hfq["low"] = temp_df_hfq["low"] * temp_df_hfq["hfq_factor"]
ret["hfq"] = temp_df_hfq.iloc[:, :-1]
# "qfq":
temp_df_qfq = temp_df.copy()
res = requests.get(zh_sina_a_stock_qfq_url.format(symbol))
qfq_factor_df = pd.DataFrame(
eval(res.text.split("=")[1].split("\n")[0])['data'])
qfq_factor_df.columns = ["date", "qfq_factor"]
qfq_factor_df.index = pd.to_datetime(qfq_factor_df.date)
del qfq_factor_df["date"]
temp_df_qfq = pd.merge(temp_df_qfq,
qfq_factor_df,
left_index=True,
right_index=True,
how="left")
temp_df_qfq.fillna(method="ffill", inplace=True)
temp_df_qfq = temp_df_qfq.astype(float)
temp_df_qfq["open"] = temp_df_qfq["open"] / temp_df_qfq["qfq_factor"]
temp_df_qfq["high"] = temp_df_qfq["high"] / temp_df_qfq["qfq_factor"]
temp_df_qfq["close"] = temp_df_qfq["close"] / temp_df_qfq["qfq_factor"]
temp_df_qfq["low"] = temp_df_qfq["low"] / temp_df_qfq["qfq_factor"]
ret["qfq"] = temp_df_qfq.iloc[:, :-1]
#"hfq-factor"
res = requests.get(zh_sina_a_stock_hfq_url.format(symbol))
hfq_factor_df = pd.DataFrame(
eval(res.text.split("=")[1].split("\n")[0])['data'])
hfq_factor_df.columns = ["date", "hfq_factor"]
hfq_factor_df.index = pd.to_datetime(hfq_factor_df.date)
del hfq_factor_df["date"]
ret["hfq-factor"] = hfq_factor_df
#"qfq-factor":
res = requests.get(zh_sina_a_stock_qfq_url.format(symbol))
qfq_factor_df = pd.DataFrame(
eval(res.text.split("=")[1].split("\n")[0])['data'])
qfq_factor_df.columns = ["date", "qfq_factor"]
qfq_factor_df.index = pd.to_datetime(qfq_factor_df.date)
del qfq_factor_df["date"]
ret["qfq-factor"] = qfq_factor_df
return ret
import akshare
akshare.stock_zh_a_daily()
import sqlite3
conn = sqlite3.connect('test.db')
print("Opened database successfully")
def get_historical_news(self, start_date=None, end_date=None, freq="day"):
if end_date is None:
end_date = datetime.datetime.now().strftime("%Y%m%d")
stock_symbol_list = self.col_basic_info.distinct("symbol")
if len(stock_symbol_list) == 0:
self.get_stock_code_info()
stock_symbol_list = self.col_basic_info.distinct("symbol")
if freq == "day":
start_stock_code = 0 if self.redis_client.get(
"start_stock_code") is None else int(
self.redis_client.get("start_stock_code").decode())
for symbol in stock_symbol_list:
if int(symbol[2:]) > start_stock_code:
if start_date is None:
# 如果该symbol有历史数据，如果有则从API获取从数据库中最近的时间开始直到现在的所有价格数据
# 如果该symbol无历史数据，则从API获取从2015年1月1日开始直到现在的所有价格数据
_latest_date = self.redis_client.get(symbol)
if _latest_date is None:
symbol_start_date = config.STOCK_PRICE_REQUEST_DEFAULT_DATE
else:
tmp_date_dt = datetime.datetime.strptime(
_latest_date.decode(), "%Y-%m-%d").date()
offset = datetime.timedelta(days=1)
symbol_start_date = (tmp_date_dt +
offset).strftime('%Y%m%d')
if symbol_start_date < end_date:
stock_zh_a_daily_hfq_df = ak.stock_zh_a_daily(
symbol=symbol,
start_date=symbol_start_date,
end_date=end_date,
adjust="qfq")
stock_zh_a_daily_hfq_df.insert(
0, 'date', stock_zh_a_daily_hfq_df.index.tolist())
stock_zh_a_daily_hfq_df.index = range(
len(stock_zh_a_daily_hfq_df))
_col = self.db_obj.get_collection(
self.database_name, symbol)
for _id in range(stock_zh_a_daily_hfq_df.shape[0]):
_tmp_dict = stock_zh_a_daily_hfq_df.iloc[
_id].to_dict()
_tmp_dict.pop("outstanding_share")
_tmp_dict.pop("turnover")
_col.insert_one(_tmp_dict)
self.redis_client.set(
symbol,
str(_tmp_dict["date"]).split(" ")[0])
logging.info(
"{} finished saving from {} to {} ... ".format(
symbol, symbol_start_date, end_date))
self.redis_client.set("start_stock_code", int(symbol[2:]))
self.redis_client.set("start_stock_code", 0)
elif freq == "week":
pass
elif freq == "month":
pass
elif freq == "5mins":
pass
elif freq == "15mins":
pass
elif freq == "30mins":
pass
elif freq == "60mins":
pass
def statisticsGet(stockCode, stockRenewed=""):
datatable = ak.stock_zh_a_daily(symbol=stockCode, adjust=stockRenewed)
datatable = datatable.dropna()
datatable = datatable.reset_index()
return datatable
def query_stock(stock_name):
stock = ak.stock_zh_a_daily(symbol=query_dis[stock_name], adjust="hfq")
return stock
def stock_zh_a_minute(symbol: str = "sh600751",
period: str = "5",
adjust: str = "",
datalen: int = 2000) -> pd.DataFrame:
"""
akshare.stock_zh_a_minute 添加bar数可选
"""
url = (
"https://quotes.sina.cn/cn/api/jsonp_v2.php/=/CN_MarketDataService.getKLineData"
)
params = {
"symbol": symbol,
"scale": period,
"datalen": datalen,
}
r = requests.get(url, params=params)
temp_df = pd.DataFrame(json.loads(
r.text.split("=(")[1].split(");")[0])).iloc[:, :6]
try:
ak.stock_zh_a_daily(symbol=symbol, adjust="qfq")
except:
return temp_df
if adjust == "":
return temp_df
if adjust == "qfq":
temp_df[["date", "time"]] = temp_df["day"].str.split(" ", expand=True)
need_df = temp_df[temp_df["time"] == "15:00:00"]
need_df.index = pd.to_datetime(need_df["date"])
stock_zh_a_daily_qfq_df = ak.stock_zh_a_daily(symbol=symbol,
adjust="qfq")
result_df = stock_zh_a_daily_qfq_df.iloc[-len(need_df):, :][
"close"].astype(float) / need_df["close"].astype(float)
temp_df.index = pd.to_datetime(temp_df["date"])
merged_df = pd.merge(temp_df,
result_df,
left_index=True,
right_index=True)
merged_df["open"] = merged_df["open"].astype(
float) * merged_df["close_y"]
merged_df["high"] = merged_df["high"].astype(
float) * merged_df["close_y"]
merged_df["low"] = merged_df["low"].astype(
float) * merged_df["close_y"]
merged_df["close"] = merged_df["close_x"].astype(
float) * merged_df["close_y"]
temp_df = merged_df[["day", "open", "high", "low", "close", "volume"]]
temp_df.reset_index(drop=True, inplace=True)
return temp_df
if adjust == "hfq":
temp_df[["date", "time"]] = temp_df["day"].str.split(" ", expand=True)
need_df = temp_df[temp_df["time"] == "15:00:00"]
need_df.index = pd.to_datetime(need_df["date"])
stock_zh_a_daily_qfq_df = ak.stock_zh_a_daily(symbol=symbol,
adjust="hfq")
result_df = stock_zh_a_daily_qfq_df.iloc[-len(need_df):, :][
"close"].astype(float) / need_df["close"].astype(float)
temp_df.index = pd.to_datetime(temp_df["date"])
merged_df = pd.merge(temp_df,
result_df,
left_index=True,
right_index=True)
merged_df["open"] = merged_df["open"].astype(
float) * merged_df["close_y"]
merged_df["high"] = merged_df["high"].astype(
float) * merged_df["close_y"]
merged_df["low"] = merged_df["low"].astype(
float) * merged_df["close_y"]
merged_df["close"] = merged_df["close_x"].astype(
float) * merged_df["close_y"]
temp_df = merged_df[["day", "open", "high", "low", "close", "volume"]]
temp_df.reset_index(drop=True, inplace=True)
return temp_df
import akshare as ak
import streamlit as st
df = ak.stock_zh_a_daily(symbol="sz000002",
start_date="20101103",
end_date="20210318",
adjust="qfq")
st.write(df)
def GetStockHistory(symbol,startdate,enddate):
symbol=removedotBysymbol(symbol)
data = ak.stock_zh_a_daily(symbol=symbol, start_date=startdate, end_date=enddate, adjust="qfq")
# stock_zh_a_daily_qfq_df = ak.stock_zh_a_daily(symbol="sz000002", start_date="20101103", end_date="20201116", adjust="qfq")
data['symbol']=symbol
return data
#%%
import pandas as pd
import akshare as ak
import tushare as ts
import matplotlib.pyplot as plt
import numpy as np
import dsw
df_1 = ts.get_k_data('601727')
df_1.set_index(['date'], inplace=True)
df_2 = ak.stock_zh_a_daily("sh601727", adjust="hfq")
plt.plot(df_1['close'].values, 'r')
plt.plot(df_2['close'].values, 'b')
plt.show()
print(np.shape(df_1))
print(np.shape(df_2))
print(df_1.head())
print(df_2.head())
#%% [markdown]
#### 我们可以看到，akshare的接口查询到的数据比tushare接口因子更多，样本量也更大，所以我选择akshare接口来训练我们的模型
#%%
df = df_2
# df.to_csv('data/init_data.csv')
# %%散点图 -成交量和股价
plt.scatter(df['volume'], df['close'])
plt.xlabel('Volume')
