def ping(ip):
__time1 = datetime.datetime.now()
api = TdxHq_API()
try:
with api.connect(ip, 7709):
api.get_security_bars(9, 0, '000001', 0, 1)
return float(datetime.datetime.now() - __time1)
except:
return datetime.timedelta(9, 9, 0)
def get_kline_data():
api = TdxHq_API()
if api.connect('119.147.212.81', 7709):
data = api.get_security_bars(
9, 0, '000001', 0,
10) # 返回普通list,五个参数分别为：category（k线),市场代码(深市),股票代码,开始时间,记录条数
print(data)
data = api.to_df(api.get_security_bars(9, 0, '000001', 0,
10)) # 返回DataFrame
print(data)
api.disconnect()
def QA_fetch_get_stock_day(code, start_date, end_date, if_fq='00', frequence='day', ip=None, port=None):
"""获取日线及以上级别的数据
Arguments:
code {str:6} -- code 是一个单独的code 6位长度的str
start_date {str:10} -- 10位长度的日期 比如'2017-01-01'
end_date {str:10} -- 10位长度的日期 比如'2018-01-01'
Keyword Arguments:
if_fq {str} -- '00'/'bfq' -- 不复权 '01'/'qfq' -- 前复权 '02'/'hfq' -- 后复权 '03'/'ddqfq' -- 定点前复权 '04'/'ddhfq' --定点后复权
frequency {str} -- day/week/month/quarter/year 也可以是简写 D/W/M/Q/Y
ip {str} -- [description] (default: None) ip可以通过select_best_ip()函数重新获取
port {int} -- [description] (default: {None})
Returns:
pd.DataFrame/None -- 返回的是dataframe,如果出错比如只获取了一天,而当天停牌,返回None
Exception:
如果出现网络问题/服务器拒绝, 会出现socket:time out 尝试再次获取/更换ip即可, 本函数不做处理
"""
ip, port = get_mainmarket_ip(ip, port)
api = TdxHq_API()
with api.connect(ip, port, time_out=0.7):
if frequence in ['day', 'd', 'D', 'DAY', 'Day']:
frequence = 9
elif frequence in ['w', 'W', 'Week', 'week']:
frequence = 5
elif frequence in ['month', 'M', 'm', 'Month']:
frequence = 6
elif frequence in ['quarter', 'Q', 'Quarter', 'q']:
frequence = 10
elif frequence in ['y', 'Y', 'year', 'Year']:
frequence = 11
start_date = str(start_date)[0:10]
today_ = datetime.date.today()
lens = QA_util_get_trade_gap(start_date, today_)
data = pd.concat([api.to_df(api.get_security_bars(frequence, _select_market_code(
code), code, (int(lens / 800) - i) * 800, 800)) for i in range(int(lens / 800) + 1)], axis=0)
# 这里的问题是: 如果只取了一天的股票,而当天停牌, 那么就直接返回None了
if len(data) < 1:
return None
data = data[data['open'] != 0]
data = data.assign(date=data['datetime'].apply(lambda x: str(x[0:10])),
code=str(code),\
date_stamp=data['datetime'].apply(lambda x: QA_util_date_stamp(str(x)[0:10])))\
.set_index('date', drop=False, inplace=False)
data = data.drop(['year', 'month', 'day', 'hour', 'minute', 'datetime'], axis=1)[start_date:end_date]
if if_fq in ['00','bfq']:
return data
else:
print('CURRENTLY NOT SUPPORT REALTIME FUQUAN')
return None
def QA_fetch_get_index_min(code, start, end, level='1min', ip=best_ip, port=7709):
'指数分钟线'
api = TdxHq_API()
type_ = ''
if str(level) in ['5', '5m', '5min', 'five']:
level, type_ = 0, '5min'
elif str(level) in ['1', '1m', '1min', 'one']:
level, type_ = 8, '1min'
elif str(level) in ['15', '15m', '15min', 'fifteen']:
level, type_ = 1, '15min'
elif str(level) in ['30', '30m', '30min', 'half']:
level, type_ = 2, '30min'
elif str(level) in ['60', '60m', '60min', '1h']:
level, type_ = 3, '60min'
with api.connect(ip, port):
if str(code)[0] in ['5', '1']: # ETF
data = pd.concat([api.to_df(api.get_security_bars(
level, 1 if str(code)[0] in ['0', '8', '9', '5'] else 0, code, (25 - i) * 800, 800)) for i in range(26)], axis=0)
else:
data = pd.concat([api.to_df(api.get_index_bars(
level, 1 if str(code)[0] in ['0', '8', '9', '5'] else 0, code, (25 - i) * 800, 800)) for i in range(26)], axis=0)
data = data\
.assign(datetime=pd.to_datetime(data['datetime']), code=str(code))\
.drop(['year', 'month', 'day', 'hour', 'minute'], axis=1, inplace=False)\
.assign(date=data['datetime'].apply(lambda x: str(x)[0:10]))\
.assign(date_stamp=data['datetime'].apply(lambda x: QA_util_date_stamp(x)))\
.assign(time_stamp=data['datetime'].apply(lambda x: QA_util_time_stamp(x)))\
.assign(type=type_).set_index('datetime', drop=False, inplace=False)[start:end]
# data
return data.assign(datetime=data['datetime'].apply(lambda x: str(x)))
def QA_fetch_get_index_min(code, start, end, level='1min', ip=best_ip, port=7709):
api = TdxHq_API()
type_ = ''
if str(level) in ['5', '5m', '5min', 'five']:
level, type_ = 0, '5min'
elif str(level) in ['1', '1m', '1min', 'one']:
level, type_ = 8, '1min'
elif str(level) in ['15', '15m', '15min', 'fifteen']:
level, type_ = 1, '15min'
elif str(level) in ['30', '30m', '30min', 'half']:
level, type_ = 2, '30min'
elif str(level) in ['60', '60m', '60min', '1h']:
level, type_ = 3, '60min'
with api.connect(ip, port):
if str(code)[0] in ['5', '1']: # ETF
data = pd.concat([api.to_df(api.get_security_bars(
level, 1 if str(code)[0] in ['0', '8', '9', '5'] else 0, code, (25 - i) * 800, 800)) for i in range(26)], axis=0)
else:
data = pd.concat([api.to_df(api.get_index_bars(
level, 1 if str(code)[0] in ['0', '8', '9', '5'] else 0, code, (25 - i) * 800, 800)) for i in range(26)], axis=0)
data = data\
.assign(datetime=pd.to_datetime(data['datetime']), code=str(code))\
.drop(['year', 'month', 'day', 'hour', 'minute'], axis=1, inplace=False)\
.assign(date=data['datetime'].apply(lambda x: str(x)[0:10]))\
.assign(date_stamp=data['datetime'].apply(lambda x: QA_util_date_stamp(x)))\
.assign(time_stamp=data['datetime'].apply(lambda x: QA_util_time_stamp(x)))\
.assign(type=type_).set_index('datetime', drop=False, inplace=False)[start:end]
# data
return data.assign(datetime=data['datetime'].apply(lambda x: str(x)))
def QA_fetch_get_index_day(code, start_date, end_date, level='day', ip=best_ip, port=7709):
'指数日线'
api = TdxHq_API()
if level in ['day', 'd', 'D', 'DAY', 'Day']:
level = 9
elif level in ['w', 'W', 'Week', 'week']:
level = 5
elif level in ['month', 'M', 'm', 'Month']:
level = 6
elif level in ['Q', 'Quarter', 'q']:
level = 10
elif level in ['y', 'Y', 'year', 'Year']:
level = 11
with api.connect(ip, port):
if str(code)[0] in ['5', '1']: # ETF
data = pd.concat([api.to_df(api.get_security_bars(
level, 1 if str(code)[0] in ['0', '8', '9', '5'] else 0, code, (25 - i) * 800, 800)) for i in range(26)], axis=0)
else:
data = pd.concat([api.to_df(api.get_index_bars(
level, 1 if str(code)[0] in ['0', '8', '9', '5'] else 0, code, (25 - i) * 800, 800)) for i in range(26)], axis=0)
data = data.assign(date=data['datetime'].apply(lambda x: str(x[0:10]))).assign(code=str(code))\
.assign(date_stamp=data['datetime'].apply(lambda x: QA_util_date_stamp(str(x)[0:10])))\
.set_index('date', drop=False, inplace=False)\
.drop(['year', 'month', 'day', 'hour',
'minute', 'datetime'], axis=1)[start_date:end_date]
return data.assign(date=data['date'].apply(lambda x: str(x)[0:10]))
def QA_fetch_get_security_bars(code, _type, lens, ip=best_ip['stock'], port=7709):
"""按bar长度推算数据
Arguments:
code {[type]} -- [description]
_type {[type]} -- [description]
lens {[type]} -- [description]
Keyword Arguments:
ip {[type]} -- [description] (default: {best_ip})
port {[type]} -- [description] (default: {7709})
Returns:
[type] -- [description]
"""
api = TdxHq_API()
with api.connect(ip, port):
data = pd.concat([api.to_df(api.get_security_bars(_select_type(_type), _select_market_code(
code), code, (i - 1) * 800, 800)) for i in range(1, int(lens / 800) + 2)], axis=0)
data = data\
.assign(datetime=pd.to_datetime(data['datetime']), code=str(code))\
.drop(['year', 'month', 'day', 'hour', 'minute'], axis=1, inplace=False)\
.assign(date=data['datetime'].apply(lambda x: str(x)[0:10]))\
.assign(date_stamp=data['datetime'].apply(lambda x: QA_util_date_stamp(x)))\
.assign(time_stamp=data['datetime'].apply(lambda x: QA_util_time_stamp(x)))\
.assign(type=_type).set_index('datetime', drop=False, inplace=False).tail(lens)
if data is not None:
return data
else:
return None
def _get_bars(self, market, code, ktype):
data = []
tdx_market = self._trans_market(market)
if tdx_market is None:
print("tdx_market == None")
return data
tdx_ktype = self._trans_ktype(ktype)
if tdx_ktype is None:
print("tdx_ktype == None")
return data
try:
ip = self.getParam('ip')
port = self.getParam('port')
except:
ip = '119.147.212.81'
port = 7709
api = TdxHq_API(raise_exception=True)
with api.connect(ip, port):
if (market == 'SH' and code[:3] == '000') \
or (market == 'SZ' and code[:2] == '39'):
for i in range(self._max[ktype]):
data += api.get_index_bars(tdx_ktype, tdx_market, code,
(self._max[ktype]-1-i)*800,800)
else:
for i in range(self._max[ktype]):
data += api.get_security_bars(tdx_ktype, tdx_market, code,
(self._max[ktype]-1-i)*800,800)
return data
def QA_fetch_get_stock_min(code, start, end, frequence='1min', ip=best_ip['stock'], port=7709):
api = TdxHq_API()
type_ = ''
start_date = str(start)[0:10]
today_ = datetime.date.today()
lens = QA_util_get_trade_gap(start_date, today_)
if str(frequence) in ['5', '5m', '5min', 'five']:
frequence, type_ = 0, '5min'
lens = 48 * lens
elif str(frequence) in ['1', '1m', '1min', 'one']:
frequence, type_ = 8, '1min'
lens = 240 * lens
elif str(frequence) in ['15', '15m', '15min', 'fifteen']:
frequence, type_ = 1, '15min'
lens = 16 * lens
elif str(frequence) in ['30', '30m', '30min', 'half']:
frequence, type_ = 2, '30min'
lens = 8 * lens
elif str(frequence) in ['60', '60m', '60min', '1h']:
frequence, type_ = 3, '60min'
lens = 4 * lens
if lens > 20800:
lens = 20800
with api.connect(ip, port):
data = pd.concat([api.to_df(api.get_security_bars(frequence, _select_market_code(
str(code)), str(code), (int(lens / 800) - i) * 800, 800)) for i in range(int(lens / 800) + 1)], axis=0)
data = data\
.assign(datetime=pd.to_datetime(data['datetime']), code=str(code))\
.drop(['year', 'month', 'day', 'hour', 'minute'], axis=1, inplace=False)\
.assign(date=data['datetime'].apply(lambda x: str(x)[0:10]))\
.assign(date_stamp=data['datetime'].apply(lambda x: QA_util_date_stamp(x)))\
.assign(time_stamp=data['datetime'].apply(lambda x: QA_util_time_stamp(x)))\
.assign(type=type_).set_index('datetime', drop=False, inplace=False)[start:end]
return data.assign(datetime=data['datetime'].apply(lambda x: str(x)))
def QA_fetch_get_index_day(code, start_date, end_date, level='day', ip=best_ip, port=7709):
'指数日线'
api = TdxHq_API()
if level in ['day', 'd', 'D', 'DAY', 'Day']:
level = 9
elif level in ['w', 'W', 'Week', 'week']:
level = 5
elif level in ['month', 'M', 'm', 'Month']:
level = 6
elif level in ['Q', 'Quarter', 'q']:
level = 10
elif level in ['y', 'Y', 'year', 'Year']:
level = 11
with api.connect(ip, port):
if str(code)[0] in ['5', '1']: # ETF
data = pd.concat([api.to_df(api.get_security_bars(
level, 1 if str(code)[0] in ['0', '8', '9', '5'] else 0, code, (25 - i) * 800, 800)) for i in range(26)], axis=0)
else:
data = pd.concat([api.to_df(api.get_index_bars(
level, 1 if str(code)[0] in ['0', '8', '9', '5'] else 0, code, (25 - i) * 800, 800)) for i in range(26)], axis=0)
data = data.assign(date=data['datetime'].apply(lambda x: str(x[0:10]))).assign(code=str(code))\
.assign(date_stamp=data['datetime'].apply(lambda x: QA_util_date_stamp(str(x)[0:10])))\
.set_index('date', drop=False, inplace=False)\
.drop(['year', 'month', 'day', 'hour',
'minute', 'datetime'], axis=1)[start_date:end_date]
return data.assign(date=data['date'].apply(lambda x: str(x)[0:10]))
def QA_fetch_get_stock_min(code,
start,
end,
level,
ip='221.231.141.60',
port=7709):
api = TdxHq_API()
market_code = __select_market_code(code)
if str(level) in ['5', '5m', '5min', 'five']:
level = 0
elif str(level) in ['1', '1m', '1min', 'one']:
level = 8
elif str(level) in ['15', '15m', '15min', 'fifteen']:
level = 1
elif str(level) in ['30', '30m', '30min', 'half']:
level = 2
elif str(level) in ['60', '60m', '60min', '1h']:
level = 3
with api.connect(ip, port):
data = []
for i in range(26):
data += api.get_security_bars(level, market_code, code,
(25 - i) * 800, 800)
data = api.to_df(data)
data['datetime'] = pd.to_datetime(data['datetime'])
data['code'] = code
data = data.set_index('datetime', drop=False)
data = data.drop(['year', 'month', 'day', 'hour', 'minute'], axis=1)
data['datetime'] = data['datetime'].apply(lambda x: str(x)[0:19])
data['date'] = data['datetime'].apply(lambda x: str(x)[0:10])
data['date_stamp'] = data['date'].apply(
lambda x: QA_util_date_stamp(x))
data['time_stamp'] = data['datetime'].apply(
lambda x: QA_util_time_stamp(x))
return data[start:end]
def QA_fetch_get_stock_min(code,
start,
end,
level,
ip='221.231.141.60',
port=7709):
api = TdxHq_API()
market_code = __select_market_code(code)
if str(level) in ['5', '5m', '5min', 'five']:
level = 0
elif str(level) in ['1', '1m', '1min', 'one']:
level = 8
elif str(level) in ['15', '15m', '15min', 'fifteen']:
level = 1
elif str(level) in ['30', '30m', '30min', 'half']:
level = 2
elif str(level) in ['60', '60m', '60min', '1h']:
level = 3
with api.connect(ip, port):
data = []
for i in range(26):
data += api.get_security_bars(level, market_code, code,
(25 - i) * 800, 800)
data = api.to_df(data)
data['datetime'] = pd.to_datetime(data['datetime'])
data['code'] = code
data = data.set_index('datetime')
return data[start:end]
def QA_fetch_get_index_day(code, start_date, end_date, frequence='day', ip=best_ip['stock']['ip'], port=best_ip['stock']['port']):
'指数日线'
api = TdxHq_API()
if frequence in ['day', 'd', 'D', 'DAY', 'Day']:
frequence = 9
elif frequence in ['w', 'W', 'Week', 'week']:
frequence = 5
elif frequence in ['month', 'M', 'm', 'Month']:
frequence = 6
elif frequence in ['Q', 'Quarter', 'q']:
frequence = 10
elif frequence in ['y', 'Y', 'year', 'Year']:
frequence = 11
with api.connect(ip, port):
start_date = str(start_date)[0:10]
today_ = datetime.date.today()
lens = QA_util_get_trade_gap(start_date, today_)
if str(code)[0] in ['5', '1']: # ETF
data = pd.concat([api.to_df(api.get_security_bars(
frequence, 1 if str(code)[0] in ['0', '8', '9', '5'] else 0, code, (int(lens / 800) - i) * 800, 800)) for i in range(int(lens / 800) + 1)], axis=0)
else:
data = pd.concat([api.to_df(api.get_index_bars(
frequence, 1 if str(code)[0] in ['0', '8', '9', '5'] else 0, code, (int(lens / 800) - i) * 800, 800)) for i in range(int(lens / 800) + 1)], axis=0)
data = data.assign(date=data['datetime'].apply(lambda x: str(x[0:10]))).assign(code=str(code))\
.assign(date_stamp=data['datetime'].apply(lambda x: QA_util_date_stamp(str(x)[0:10])))\
.set_index('date', drop=False, inplace=False)\
.assign(code=code)\
.drop(['year', 'month', 'day', 'hour',
'minute', 'datetime'], axis=1)[start_date:end_date]
return data.assign(date=data['date'].apply(lambda x: str(x)[0:10]))
def fetch_get_stock_min(code, start, end, level='1min', ip=ac.TDX_BEST_IP, port=7709):
api = TdxHq_API()
type_ = ''
if str(level) in ['5', '5m', '5min', 'five']:
level, type_ = 0, '5min'
elif str(level) in ['1', '1m', '1min', 'one']:
level, type_ = 8, '1min'
elif str(level) in ['15', '15m', '15min', 'fifteen']:
level, type_ = 1, '15min'
elif str(level) in ['30', '30m', '30min', 'half']:
level, type_ = 2, '30min'
elif str(level) in ['60', '60m', '60min', '1h']:
level, type_ = 3, '60min'
with api.connect(ip, port):
data = pd.concat([api.to_df(api.get_security_bars(level, __select_market_code(
str(code)), str(code), (25 - i) * 800, 800)) for i in range(26)], axis=0)
data = data\
.assign(datetime=pd.to_datetime(data['datetime']), code=str(code))\
.drop(['year', 'month', 'day', 'hour', 'minute'], axis=1, inplace=False)\
.assign(date=data['datetime'].apply(lambda x: str(x)[0:10]))\
.assign(date_stamp=data['datetime'].apply(lambda x: util_date_stamp(x)))\
.assign(time_stamp=data['datetime'].apply(lambda x: util_time_stamp(x)))\
.assign(type=type_).set_index('datetime', drop=False, inplace=False)[start:end]
return data.assign(datetime=data['datetime'].apply(lambda x: str(x)))
def single_task(code, timeout=100):
api = TdxHq_API()
api.connect('115.238.90.165', 7709)
market = __select_market_code(code)
res = api.get_security_bars(1, market, code, 0, 1)[0]
#re=[change(x,code) for x in res]
#res=api.get_security_quotes([(__select_market_code(code), code)])
return changer(res, code)
def fetch_security_bars(code, _type, lens, ip=ac.TDX_BEST_IP, port=7709):
api = TdxHq_API()
with api.connect(ip, port):
data = pd.concat([api.to_df(api.get_security_bars(__select_type(_type), __select_market_code(
code), code, (i - 1) * 800, 800)) for i in range(1, int(lens / 800) + 2)], axis=0)
if data is not None:
return data
else:
return None
def multiple_task(code, timeout=100):
api = TdxHq_API()
api.connect('115.238.90.165', 7709)
market = __select_market_code(code)
res = api.get_security_bars(1, market, code, 0, 800)
re=[changer(x,code) for x in res]
#print(re)
#res=api.get_security_quotes([(__select_market_code(code), code)])
return re
def QA_fetch_get_index_min(code, start, end, frequence='1min', ip=None, port=None):
'指数分钟线'
global best_ip
if ip is None and port is None and best_ip['stock']['ip'] is None and best_ip['stock']['port'] is None:
best_ip = select_best_ip()
ip = best_ip['stock']['ip']
port = best_ip['stock']['port']
elif ip is None and port is None and best_ip['stock']['ip'] is not None and best_ip['stock']['port'] is not None:
ip = best_ip['stock']['ip']
port = best_ip['stock']['port']
else:
pass
api = TdxHq_API()
type_ = ''
start_date = str(start)[0:10]
today_ = datetime.date.today()
lens = QA_util_get_trade_gap(start_date, today_)
if str(frequence) in ['5', '5m', '5min', 'five']:
frequence, type_ = 0, '5min'
lens = 48 * lens
elif str(frequence) in ['1', '1m', '1min', 'one']:
frequence, type_ = 8, '1min'
lens = 240 * lens
elif str(frequence) in ['15', '15m', '15min', 'fifteen']:
frequence, type_ = 1, '15min'
lens = 16 * lens
elif str(frequence) in ['30', '30m', '30min', 'half']:
frequence, type_ = 2, '30min'
lens = 8 * lens
elif str(frequence) in ['60', '60m', '60min', '1h']:
frequence, type_ = 3, '60min'
lens = 4 * lens
if lens > 20800:
lens = 20800
with api.connect(ip, port):
if str(code)[0] in ['5', '1']: # ETF
data = pd.concat([api.to_df(api.get_security_bars(
frequence, 1 if str(code)[0] in ['0', '8', '9', '5'] else 0, code, (int(lens / 800) - i) * 800, 800)) for i in range(int(lens / 800) + 1)], axis=0)
else:
data = pd.concat([api.to_df(api.get_index_bars(
frequence, 1 if str(code)[0] in ['0', '8', '9', '5'] else 0, code, (int(lens / 800) - i) * 800, 800)) for i in range(int(lens / 800) + 1)], axis=0)
data = data\
.assign(datetime=pd.to_datetime(data['datetime']), code=str(code))\
.drop(['year', 'month', 'day', 'hour', 'minute'], axis=1, inplace=False)\
.assign(code=code)\
.assign(date=data['datetime'].apply(lambda x: str(x)[0:10]))\
.assign(date_stamp=data['datetime'].apply(lambda x: QA_util_date_stamp(x)))\
.assign(time_stamp=data['datetime'].apply(lambda x: QA_util_time_stamp(x)))\
.assign(type=type_).set_index('datetime', drop=False, inplace=False)[start:end]
# data
return data.assign(datetime=data['datetime'].apply(lambda x: str(x)))
def QA_fetch_get_stock_latest(code, ip=best_ip, port=7709):
code = [code] if isinstance(code, str) else code
api = TdxHq_API(multithread=True)
with api.connect(ip, port):
data = pd.concat([api.to_df(api.get_security_bars(
9, __select_market_code(item), item, 0, 1)).assign(code=item) for item in code], axis=0)
return data\
.assign(date=pd.to_datetime(data['datetime']
.apply(lambda x: x[0:10])), date_stamp=data['datetime']
.apply(lambda x: QA_util_date_stamp(str(x[0:10]))))\
.set_index('date', drop=False)\
.drop(['year', 'month', 'day', 'hour', 'minute', 'datetime'], axis=1)
def QA_fetch_get_stock_latest(code, ip=best_ip, port=7709):
code = [code] if isinstance(code, str) else code
api = TdxHq_API(multithread=True)
with api.connect(ip, port):
data = pd.concat([api.to_df(api.get_security_bars(
9, __select_market_code(item), item, 0, 1)).assign(code=item) for item in code], axis=0)
return data\
.assign(date=pd.to_datetime(data['datetime']
.apply(lambda x: x[0:10])), date_stamp=data['datetime']
.apply(lambda x: QA_util_date_stamp(str(x[0:10]))))\
.set_index('date', drop=False)\
.drop(['year', 'month', 'day', 'hour', 'minute', 'datetime'], axis=1)
def __get_trade_hq(self, code, days=10, adj='qfq'):
mkcode = 1 if code.startswith('6') else 0
api = TdxHq_API()
with api.connect('218.108.98.244', 7709):
# 获取股票行情数据，不复权的数据
data = api.get_security_bars(TDXParams.KLINE_TYPE_DAILY, mkcode,
code, 0, days + 1)
data = api.to_df(data)
data['datetime'] = data['datetime'].apply(lambda x: str(x[0:10]))
data['datetime'] = pd.to_datetime(data['datetime'])
data = data.assign(code=str(code)).set_index(
'datetime', drop=True,
inplace=False).drop(['year', 'month', 'day', 'hour', 'minute'],
axis=1)
data = data.sort_index(ascending=False)
# 加入复权因子，计算复权数据
if adj:
df_adj = pd.read_csv(
'http://file.tushare.org/tsdata/f/factor/{0}.csv'.format(
code))
df_adj = df_adj.set_index('datetime')
data = data.merge(df_adj,
how='left',
left_index=True,
right_index=True)
data['adj_factor'] = data['adj_factor'].fillna(method='bfill')
for col in ['open', 'close', 'high', 'low']:
if adj == 'qfq':
data[col] = data[col] * data['adj_factor'] / float(
df_adj['adj_factor'][0])
else:
data[col] = data[col] * data[
'adj_factor'] # data[col] = data[col].map(lambda x: '%.2f' % x) # 会引起类型转换，从float转换成str
# 计算换手率
cursor = connection.cursor()
cursor.execute(
'select outstanding from stock_stockbasicinfo where code="{0}"'
.format(code))
outstanding = cursor.fetchone()[0] * 10**8
data['outstanding'] = outstanding
data['turnover'] = data['vol'] / data['outstanding'] * 100
# 计算涨跌幅
data['pre_close'] = data['close'].shift(-1)
data['change_rate'] = (data['close'] -
data['pre_close']) / data['pre_close'] * 100
data = data.dropna()
data = data[[
'code', 'pre_close', 'open', 'close', 'high', 'low', 'vol',
'amount', 'change_rate', 'turnover', 'outstanding'
]]
return data
def QA_fetch_get_index_day(code, start_date, end_date, frequence='day', ip=None, port=None):
"""指数日线
1- sh
0 -sz
Arguments:
code {[type]} -- [description]
start_date {[type]} -- [description]
end_date {[type]} -- [description]
Keyword Arguments:
frequence {str} -- [description] (default: {'day'})
ip {[type]} -- [description] (default: {None})
port {[type]} -- [description] (default: {None})
Returns:
[type] -- [description]
"""
ip, port = get_mainmarket_ip(ip, port)
api = TdxHq_API()
if frequence in ['day', 'd', 'D', 'DAY', 'Day']:
frequence = 9
elif frequence in ['w', 'W', 'Week', 'week']:
frequence = 5
elif frequence in ['month', 'M', 'm', 'Month']:
frequence = 6
elif frequence in ['Q', 'Quarter', 'q']:
frequence = 10
elif frequence in ['y', 'Y', 'year', 'Year']:
frequence = 11
with api.connect(ip, port):
start_date = str(start_date)[0:10]
today_ = datetime.date.today()
lens = QA_util_get_trade_gap(start_date, today_)
if str(code)[0] in ['5', '1']: # ETF
data = pd.concat([api.to_df(api.get_security_bars(
frequence, 1 if str(code)[0] in ['0', '8', '9', '5'] else 0, code, (int(lens / 800) - i) * 800, 800)) for i in range(int(lens / 800) + 1)], axis=0)
else:
data = pd.concat([api.to_df(api.get_index_bars(
frequence, 1 if str(code)[0] in ['0', '8', '9', '5'] else 0, code, (int(lens / 800) - i) * 800, 800)) for i in range(int(lens / 800) + 1)], axis=0)
data = data.assign(date=data['datetime'].apply(lambda x: str(x[0:10]))).assign(code=str(code))\
.assign(date_stamp=data['datetime'].apply(lambda x: QA_util_date_stamp(str(x)[0:10])))\
.set_index('date', drop=False, inplace=False)\
.assign(code=code)\
.drop(['year', 'month', 'day', 'hour',
'minute', 'datetime'], axis=1)[start_date:end_date]
return data.assign(date=data['date'].apply(lambda x: str(x)[0:10]))
def ping(ip, port=7709, multithread=False):
api = TdxHq_API(multithread=multithread)
success = False
starttime = time.time()
try:
with api.connect(ip, port, time_out=1):
#x = api.get_security_count(0)
#x = api.get_index_bars(7, 1, '000001', 800, 100)
x = api.get_security_bars(7, 0, '000001', 800, 100)
if x:
success = True
except Exception as e:
success = False
endtime = time.time()
return (success, endtime - starttime, ip, port)
def ping(ip, port=7709, multithread=False):
api = TdxHq_API(multithread=multithread)
success = False
starttime = time.time()
try:
with api.connect(ip, port, time_out=1):
#x = api.get_security_count(0)
#x = api.get_index_bars(7, 1, '000001', 800, 100)
x = api.get_security_bars(7, 0, '000001', 800, 100)
if x:
success = True
except Exception as e:
success = False
endtime = time.time()
return (success, endtime - starttime, ip, port)
def QA_fetch_get_index_day(code, start_date, end_date, frequence='day', ip=None, port=None):
'指数日线'
global best_ip
if ip is None and port is None and best_ip['stock']['ip'] is None and best_ip['stock']['port'] is None:
best_ip = select_best_ip()
ip = best_ip['stock']['ip']
port = best_ip['stock']['port']
elif ip is None and port is None and best_ip['stock']['ip'] is not None and best_ip['stock']['port'] is not None:
ip = best_ip['stock']['ip']
port = best_ip['stock']['port']
else:
pass
api = TdxHq_API()
if frequence in ['day', 'd', 'D', 'DAY', 'Day']:
frequence = 9
elif frequence in ['w', 'W', 'Week', 'week']:
frequence = 5
elif frequence in ['month', 'M', 'm', 'Month']:
frequence = 6
elif frequence in ['Q', 'Quarter', 'q']:
frequence = 10
elif frequence in ['y', 'Y', 'year', 'Year']:
frequence = 11
with api.connect(ip, port):
start_date = str(start_date)[0:10]
today_ = datetime.date.today()
lens = QA_util_get_trade_gap(start_date, today_)
if str(code)[0] in ['5', '1']: # ETF
data = pd.concat([api.to_df(api.get_security_bars(
frequence, 1 if str(code)[0] in ['0', '8', '9', '5'] else 0, code, (int(lens / 800) - i) * 800, 800)) for i in range(int(lens / 800) + 1)], axis=0)
else:
data = pd.concat([api.to_df(api.get_index_bars(
frequence, 1 if str(code)[0] in ['0', '8', '9', '5'] else 0, code, (int(lens / 800) - i) * 800, 800)) for i in range(int(lens / 800) + 1)], axis=0)
data = data.assign(date=data['datetime'].apply(lambda x: str(x[0:10]))).assign(code=str(code))\
.assign(date_stamp=data['datetime'].apply(lambda x: QA_util_date_stamp(str(x)[0:10])))\
.set_index('date', drop=False, inplace=False)\
.assign(code=code)\
.drop(['year', 'month', 'day', 'hour',
'minute', 'datetime'], axis=1)[start_date:end_date]
return data.assign(date=data['date'].apply(lambda x: str(x)[0:10]))
def QA_fetch_get_stock_day(code,
start_date,
end_date,
ip='119.147.212.81',
port=7709):
api = TdxHq_API()
market_code = __select_market_code(code)
start_date = QA_util_get_real_date(start_date, trade_date_sse, 1)
end_date = QA_util_get_real_date(end_date, trade_date_sse, -1)
with api.connect(ip, port):
data = []
for i in range(10):
data += api.get_security_bars(9, market_code, code, (9 - i) * 800,
800)
data = api.to_df(data)
data['date'] = data['datetime'].apply(lambda x: x[0:10])
data['date'] = pd.to_datetime(data['date'])
data = data.set_index('date')
data = data.drop(
['year', 'month', 'day', 'hour', 'minute', 'datetime'], axis=1)
return data[start_date:end_date]
def download_stock_bars(stock_code, freq='day', fq_type='qfq'):
"""获取股票"""
if stock_code[0] == '0':
market = 0
else:
market = 1
category = FREQUENT[freq]
api = TdxHq_API()
with api.connect('119.147.212.81', 7709):
index = 0
stock_data = []
while True:
older_data = api.get_security_bars(category, market, stock_code, index * 800, 800)
if older_data:
stock_data = older_data + stock_data
index += 1
else:
break
stock_df = api.to_df(stock_data) # 返回DataFrame
return stock_df[ct.STOCK_COLS]
def QA_fetch_get_security_bars(code, _type, lens, ip=None, port=None):
"""按bar长度推算数据
Arguments:
code {[type]} -- [description]
_type {[type]} -- [description]
lens {[type]} -- [description]
Keyword Arguments:
ip {[type]} -- [description] (default: {best_ip})
port {[type]} -- [description] (default: {7709})
Returns:
[type] -- [description]
"""
global best_ip
if ip is None and port is None and best_ip['stock']['ip'] is None and best_ip['stock']['port'] is None:
best_ip = select_best_ip()
ip = best_ip['stock']['ip']
port = best_ip['stock']['port']
elif ip is None and port is None and best_ip['stock']['ip'] is not None and best_ip['stock']['port'] is not None:
ip = best_ip['stock']['ip']
port = best_ip['stock']['port']
else:
pass
api = TdxHq_API()
with api.connect(ip, port):
data = pd.concat([api.to_df(api.get_security_bars(_select_type(_type), _select_market_code(
code), code, (i - 1) * 800, 800)) for i in range(1, int(lens / 800) + 2)], axis=0)
data = data\
.assign(datetime=pd.to_datetime(data['datetime']), code=str(code))\
.drop(['year', 'month', 'day', 'hour', 'minute'], axis=1, inplace=False)\
.assign(date=data['datetime'].apply(lambda x: str(x)[0:10]))\
.assign(date_stamp=data['datetime'].apply(lambda x: QA_util_date_stamp(x)))\
.assign(time_stamp=data['datetime'].apply(lambda x: QA_util_time_stamp(x)))\
.assign(type=_type).set_index('datetime', drop=False, inplace=False).tail(lens)
if data is not None:
return data
else:
return None
def QA_fetch_get_stock_latest(code, ip=None, port=None):
global best_ip
if ip is None and port is None and best_ip['stock']['ip'] is None and best_ip['stock']['port'] is None:
best_ip = select_best_ip()
ip = best_ip['stock']['ip']
port = best_ip['stock']['port']
elif ip is None and port is None and best_ip['stock']['ip'] is not None and best_ip['stock']['port'] is not None:
ip = best_ip['stock']['ip']
port = best_ip['stock']['port']
else:
pass
code = [code] if isinstance(code, str) else code
api = TdxHq_API(multithread=True)
with api.connect(ip, port):
data = pd.concat([api.to_df(api.get_security_bars(
9, _select_market_code(item), item, 0, 1)).assign(code=item) for item in code], axis=0)
return data\
.assign(date=pd.to_datetime(data['datetime']
.apply(lambda x: x[0:10])), date_stamp=data['datetime']
.apply(lambda x: QA_util_date_stamp(str(x[0:10]))))\
.set_index('date', drop=False)\
.drop(['year', 'month', 'day', 'hour', 'minute', 'datetime'], axis=1)
def test_api():
ip = '14.17.75.71'
api = TdxHq_API()
api.connect(ip)
freq = 8
exchange = 0
code = '000001'
SECURITY_BARS_PATCH_SIZE = 800
SECURITY_BARS_PATCH_NUM2 = 30
now = datetime.datetime.now()
for i in range(3):
dd = [
api.get_security_bars(freq, exchange, code,
(SECURITY_BARS_PATCH_NUM2 - i - 1) *
SECURITY_BARS_PATCH_SIZE,
SECURITY_BARS_PATCH_SIZE)
for i in range(SECURITY_BARS_PATCH_NUM2)
]
print((datetime.datetime.now() - now).total_seconds())
class TdxData:
def __init__(self):
self.api = TdxHq_API(heartbeat=True)
self.dbUtil = DBUtil.getInstance()
def get_security_quotes(self, code, type):
return self.api.get_security_quotes([(type, code)])
# 支持板块及个股
def days(self, code, type, bk=False, all=False, day=5):
category = int(config.getByKey('TDX_CATEGORY'))
try:
with self.api.connect(TDX_IP, TDX_PORT):
data = []
if all:
if bk:
for i in range(10):
data += self.api.get_index_bars(
category, type, code, (9 - i) * 800, 800)
else:
for i in range(10):
data += self.api.get_security_bars(
category, type, code, (9 - i) * 800, 800)
if len(data) > 0:
df = self.api.to_df(data).drop([
'amount', 'year', 'month', 'day', 'hour', 'minute'
],
axis=1)
df['trade_date'] = df['datetime'].apply(
lambda x: x[0:10].replace('-', ''))
df = df.drop(['datetime'], axis=1)
df = df.sort_values(by=['trade_date'],
axis=0,
ascending=False)
return df
else:
return self.api.to_df(data)
else:
if bk:
data = self.api.get_index_bars(category, type, code, 0,
day) # 返回DataFrame
else:
data = self.api.get_security_bars(
category, type, code, 0, day)
if len(data) > 0:
df = self.api.to_df(data).drop([
'amount', 'year', 'month', 'day', 'hour', 'minute'
],
axis=1)
df['trade_date'] = df['datetime'].apply(
lambda x: x[0:10].replace('-', ''))
df = df.drop(['datetime'], axis=1)
df = df.sort_values(by=['trade_date'],
axis=0,
ascending=False)
return df
else:
return self.api.to_df(data)
except Exception as e:
logging.info("暂不支持类型,代码：%s:%s" % (code, e))
return self.api.to_df([])
# F10 查询公司信息目录
def get_company_info_category(self, code, type):
with self.api.connect(TDX_IP, TDX_PORT):
df = pd.DataFrame(self.api.get_company_info_category(type, code))
df['txt'] = None
return df
return []
def get_company_info_content(self, code, type, df):
with self.api.connect(TDX_IP, TDX_PORT):
return self.api.get_company_info_content(type, code,
df['filename'].values[0],
df['start'].values[0],
df['length'].values[0])
return ""
# 查询财务数据
def get_finance_info(self, code, type):
with self.api.connect(TDX_IP, TDX_PORT):
return self.api.get_finance_info(type, code)
return ''
# 每年更新一次，板块个股关系
def updateBk(self):
with self.api.connect(TDX_IP, TDX_PORT):
"""
# 获取股票所属板块信息
# 板块相关参数
BLOCK_SZ = "block_zs.dat"
BLOCK_FG = "block_fg.dat"
BLOCK_GN = "block_gn.dat"
BLOCK_DEFAULT = "block.dat"
"""
bk_zs = self.api.to_df(
self.api.get_and_parse_block_info("block_zs.dat")) #指数板块
bk_fg = self.api.to_df(
self.api.get_and_parse_block_info("block_fg.dat")) #风格板块
bk_gn = self.api.to_df(
self.api.get_and_parse_block_info("block_gn.dat")) #概念板块
bk_default = self.api.to_df(
self.api.get_and_parse_block_info("block.dat")) # 默认
self.dbUtil.to_sql(bk_gn,
'stock_bk_gn',
if_exists='replace',
index=False,
dtype=STOCK_BK_DTYPE)
self.dbUtil.to_sql(bk_fg,
'stock_bk_fg',
if_exists='replace',
index=False,
dtype=STOCK_BK_DTYPE)
self.dbUtil.to_sql(bk_zs,
'stock_bk_zs',
if_exists='replace',
index=False,
dtype=STOCK_BK_DTYPE)
self.dbUtil.to_sql(bk_default,
'stock_bk_default',
if_exists='replace',
index=False,
dtype=STOCK_BK_DTYPE)
# 获取股票列表
tmp1 = self.api.to_df(self.api.get_security_list(0, 0)) # 深圳
tmp1['type'] = 0
tmp2 = self.api.to_df(self.api.get_security_list(1, 0)) # 上海
tmp2['type'] = 1
tmp = tmp1.append(tmp2)
self.dbUtil.to_sql(tmp,
'stock_bk',
if_exists='replace',
index=False,
dtype=STOCK_BK)
def updateGD(self, code, type):
url = 'http://emweb.securities.eastmoney.com/PC_HSF10/ShareholderResearch/ShareholderResearchAjax?code=%s%s' % (
type.lower(), code)
html = urllib.request.urlopen(url).read()
# 将字符串转换成字典
data = json.loads(html.decode('utf-8'))
# gdrs 股东人数,sdgd 十大股东 ，sdltgd 十大流通股东
df_gdrs = pd.DataFrame(data['gdrs'])
df_gdrs['code'] = code
try:
db_df_gdrs = self.dbUtil.read_sql(
"select * from stock_gdrs where code ='%s'" % code)
# 数据合并
df_gdrs = df_gdrs.append(db_df_gdrs).drop_duplicates(
subset=['code', 'rq', 'gdmc'], keep='last')
except Exception as e:
pass
self.dbUtil.to_sql(df_gdrs,
'stock_gdrs',
if_exists='append',
index=False,
dtype=STOCK_GDRS_DTYPE)
sdgd = []
for i in range(len(data['sdgd'])):
sdgd += data['sdgd'][i]['sdgd']
df_sdgd = pd.DataFrame(sdgd)
df_sdgd['code'] = code
try:
db_df_sdgd = self.dbUtil.read_sql(
"select * from stock_sdgd where code ='%s'" % code)
df_sdgd = df_sdgd.append(db_df_sdgd).drop_duplicates(
subset=['code', 'rq', 'gdmc'], keep='last')
except Exception as e:
pass
self.dbUtil.to_sql(df_sdgd,
'stock_sdgd',
if_exists='append',
index=False,
dtype=STOCK_SDGD_DTYPE)
sdltgd = []
for i in range(len(data['sdltgd'])):
sdltgd += data['sdltgd'][i]['sdltgd']
df_sdltgd = pd.DataFrame(sdltgd)
df_sdltgd['code'] = code
# 获取后与数据库中的数据进行merge,首次表不存在，会抛异常
try:
db_df_sdltgd = self.dbUtil.read_sql(
"select * from stock_sdltgd where code ='%s'" % code)
df_sdltgd = df_sdltgd.append(db_df_sdltgd).drop_duplicates(
subset=['code', 'rq', 'gdmc'], keep='last')
except Exception as e:
pass
self.dbUtil.to_sql(df_sdltgd,
'stock_sdltgd',
if_exists='append',
index=False,
dtype=STOCK_SDGD_DTYPE)
# 没季度更新一次
def updateGDs(self):
codes = self.dbUtil.read_sql("select ts_code from stock_basic")
tmp = codes['ts_code'].str.split('.', expand=True)
for index, row in tmp.iterrows():
try:
self.updateGD(row[0], row[1])
logging.info('%s更新结束,当前索引%s' % (row[0], index))
except Exception as e:
logging.info('%s更新失败,当前索引%s' % (row[0], index))
# 分红
# 分红地址http://data.eastmoney.com/yjfp/201812.html
def updateFh(self, rq):
url = 'http://data.eastmoney.com/DataCenter_V3/yjfp/getlist.ashx?filter=(ReportingPeriod=^%s^)' % rq
html = requests.get(url)
# 将字符串转换成字典
data = json.loads(html.text)['data']
if len(data) == 0:
return 0
df = pd.DataFrame(data)
df['ReportingPeriod'] = df['ReportingPeriod'].apply(lambda x: x[0:10])
# 首次需要将df_fh制空，因为表还不存在
if self.dbUtil.is_exist("stock_fh"):
db_fh = self.dbUtil.read_sql(
"select * from stock_fh where ReportingPeriod = '%s'" %
df['ReportingPeriod'][0])
if db_fh.empty: # 不存在当前日期的分红信息，进行拼接
self.dbUtil.to_sql(df,
'stock_fh',
if_exists='append',
index=False)
return 1
else:
pass
else:
self.dbUtil.to_sql(df, 'stock_fh', if_exists='append', index=False)
return 1
# 更新历年分红
def updateFhYears(self):
now = int(time.strftime("%Y", time.localtime())) + 1
lastYear = int(self._getFhMaxYear())
for i in range(lastYear, now): #初始化时开启
type = self.updateFh('%s-06-30' % i)
logging.info('%s-06-30%s' % (i, '成功' if type == 1 else '失败'))
self.updateFh('%s-12-31' % i)
logging.info('%s-12-31%s' % (i, '成功' if type == 1 else '失败'))
def _getFhMaxYear(self):
if self.dbUtil.is_exist('stock_fh'):
try:
df = self.dbUtil.read_sql(
'select substr(max(t.ReportingPeriod ),0,5) year from stock_fh t'
)
return df['year'][0].values
except Exception as e:
pass
return 1991
@classmethod
def getInstance(cls):
if not hasattr(TdxData, "_instance"):
TdxData._instance = TdxData()
return TdxData._instance
class TDX(object):
'''
This class is tong da xin data source.
We can use it to get down the stock datas.
Tushare can't get minter line and or year line.
TDX can search index of stock and funds.
'''
def __init__(self):
self.tdx_api = TdxHq_API()
self.__ip = '119.147.212.81' #输入IP
self.__port = 7709 #端口
self.__code = '600200'
self.__market = 1 #市场代码 0:深圳，1:上海
self._startdate = "2017-01-01"
self.today = datetime.date.today()
self._enddate = datetime.datetime.strftime(self.today, '%Y-%m-%d')
self.__mkt_segment = {
'sh': '60',
"sz": '00',
"cyb": "30",
} #segment 当前板块开始字符串
def __str__(self):
return 'TDX object (code : %s)' % self.code
@property
def IP(self): # self.IP
return self.__ip
@property
def PORT(self):
return self.__port
@property
def code(self): #定义stock code 属性
return self.__code
@code.setter #设定code
def code(self, code_input):
"""
The setter of the code property
"""
if not isinstance(code_input, str): #确定是否是字符串
raise ValueError('the code must string!')
if not len(code_input) == 6: #确定长度
raise ValueError('the code value error,the len must SIX !')
if code_input.startswith('60'): #确定表头
self.__market = 1
elif code_input.startswith('00'):
self.__market = 0
elif code_input.startswith('30'):
self.__market = 0
else:
raise ValueError('this code is not stock code')
self.__code = code_input
@property
def startdate(self): #开始日期
return self._startdate
@startdate.setter #设置日期
def startdate(self, date_input):
"""
The setter of the start date property
"""
if not isinstance(date_input, str):
raise ValueError('the date must string!')
if not len(date_input) == 8:
raise ValueError(
'the date value error,the date formet must xxxx-xx-xx !')
self._startdate = date_input
@property #结束日期
def enddate(self):
return self._enddate
@enddate.setter
def enddate(self, date_input):
"""
The setter of the start date property
"""
if not isinstance(date_input, str):
raise ValueError('the date must string!')
if not len(date_input) == 8:
raise ValueError(
'the date value error,the date formet must xxxx-xx-xx !')
self._enddate = date_input
def get_day_data_tdx(self): #获取K line
with self.tdx_api.connect(self.IP, self.PORT):
data = self.tdx_api.get_k_data(self.code, self.startdate,
self.enddate)
data = pandas.DataFrame(data)
data.date = data.date.apply(
lambda x: datetime.datetime.strptime(x, "%Y-%m-%d"))
return data
#TODO: 现在是用800点进行计数，以后会细化功能
def get_k_data_tdx(self, k_mode=9):
"""
获取k 线图，总计800 点
Parameters
----------
k_mode= 0-11
0 5分钟K线
1 15分钟K线
2 30分钟K线
3 1小时K线
4 日K线
5 周K线
6 月K线
7 1分钟
8 1分钟K线 9 日K线
10 季K线
11 年K线
Returns
-------
"""
with self.tdx_api.connect(self.self.IP, self.self.PORT):
data = self.tdx_api.get_security_bars(k_mode, self.__market,
self.code, 0, 800)
data = pandas.DataFrame(data)
#data.date = data.date.apply(
# lambda x: datetime.datetime.strptime(x, "%Y-%m-%d"))
return data
def len_market(self): #市场有多少只股票
with self.tdx_api.connect(self.IP, self.PORT):
_len = self.tdx_api.get_security_count(self.__market)
return _len
def get_page_tdx(self, block=None):
if block is None:
market = 1
page = [0]
elif block in ['sh', 'SH']:
market = 1
page = [13, 14]
elif block in ['sz', 'SZ']:
print('block for shenzhen')
market = 0
page = [0, 1]
elif block in ['cyb', 'CYB']:
print('block for chuang ye ban')
market = 0
page = [7, 8]
else:
pass
code_list_df = pandas.DataFrame()
with self.tdx_api.connect(self.IP, self.PORT):
for pn in page:
data = self.tdx_api.get_security_list(market, pn * 1000)
data = pandas.DataFrame(data)
print(data)
code_list_df = code_list_df.append(data, ignore_index=True)
return code_list_df
def get_base_finace_tdx(self):
with self.tdx_api.connect(self.IP, self.PORT):
data = self.tdx_api.get_finance_info(0, '000001')
data = pandas.Series(data)
print(data)
def get_min_data(self):
from pytdx.params import TDXParams
with self.tdx_api.connect(self.IP, self.PORT):
data = self.tdx_api.get_history_minute_time_data(
TDXParams.MARKET_SH, self.code, 20161209)
data = pandas.DataFrame(data)
print(data)
#TODO: 需要确定 0: buy 1 : sell
def get_tick_data(self):
"""
历史分笔交易：time 顺序； price ; vol ;buyorsell [1:] [0:];
sh 60 13000-14000
Parameters
----------
Returns
-------
"""
data = pandas.DataFrame()
with self.tdx_api.connect(self.IP, self.PORT):
for i in [2000, 0000]:
df = self.tdx_api.get_history_transaction_data(
TDXParams.MARKET_SH, "600547", i, 2000, 20160308)
df = pandas.DataFrame(df)
data = data.append(df, ignore_index=True)
return data
def get_tick_today(self):
"""
Get every time the each deal for today.每组数最大len 2 k 所以要确定的数据长度
Parameters
----------
self:
Returns
-------
"""
with self.tdx_api.connect(self.IP, self.PORT):
data = pandas.DataFrame()
for i in [0, 2000]:
df = self.tdx_api.get_transaction_data(self.__market,
self.code, i, 2000)
df = pandas.DataFrame(df)
data = data.append(df, ignore_index=True)
return data
def get_block(self):
with self.tdx_api.connect(self.IP, self.PORT):
data = self.tdx_api.get_and_parse_block_info("block.dat")
data = pandas.DataFrame(data)
print(data)
def get_market_segment_list(self, mkt):
data = self.get_page_tdx(mkt)
self.code_list = pandas.DataFrame()
pbar = tqdm(total=len(data.code))
mkt_hard = self.mkt_segment[mkt]
for idx, __code in enumerate(data.code):
pbar.update(1)
if __code.startswith(mkt_hard, 0, 2):
self.code_list = self.code_list.append(data.loc[idx],
ignore_index=True)
return self.code_list
def get_sh_list(self):
return self.get_market_segment_list('sh')
def get_sz_list(self):
return self.get_market_segment_list('sz')
def get_cyb_list(self):
return self.get_market_segment_list('cyb')
def QA_fetch_get_stock_day(code, start_date, end_date, if_fq='00', level='day', ip=best_ip, port=7709):
api = TdxHq_API()
with api.connect(ip, port):
if level in ['day', 'd', 'D', 'DAY', 'Day']:
level = 9
elif level in ['w', 'W', 'Week', 'week']:
level = 5
elif level in ['month', 'M', 'm', 'Month']:
level = 6
elif level in ['Q', 'Quarter', 'q']:
level = 10
elif level in ['y', 'Y', 'year', 'Year']:
level = 11
data = pd.concat([api.to_df(api.get_security_bars(level, __select_market_code(
code), code, (9 - i) * 800, 800)) for i in range(10)], axis=0)
if if_fq in ['00', 'bfq']:
data = data.assign(date=data['datetime'].apply(lambda x: str(x[0:10]))).assign(code=str(code))\
.assign(date_stamp=data['datetime'].apply(lambda x: QA_util_date_stamp(str(x)[0:10])))\
.set_index('date', drop=False, inplace=False)\
.drop(['year', 'month', 'day', 'hour',
'minute', 'datetime'], axis=1)[start_date:end_date]
return data.assign(date=data['date'].apply(lambda x: str(x)[0:10]))
elif if_fq in ['01', 'qfq']:
xdxr_data = QA_fetch_get_stock_xdxr(code)
info = xdxr_data[xdxr_data['category'] == 1]
bfq_data = data\
.assign(date=pd.to_datetime(data['datetime'].apply(lambda x: x[0:10])))\
.assign(code=str(code))\
.assign(date_stamp=data['datetime'].apply(lambda x: QA_util_date_stamp(str(x)[0:10])))\
.set_index('date', drop=False, inplace=False)\
.drop(['year', 'month', 'day', 'hour',
'minute', 'datetime'], axis=1)
bfq_data['if_trade']=True
data = pd.concat([bfq_data, info[['category']]
[bfq_data.index[0]:]], axis=1)
data['date'] = data.index
data['if_trade'].fillna(value=False, inplace=True)
data = data.fillna(method='ffill')
data = pd.concat([data, info[['fenhong', 'peigu', 'peigujia',
'songzhuangu']][bfq_data.index[0]:]], axis=1)
data = data.fillna(0)
data['preclose'] = (data['close'].shift(1) * 10 - data['fenhong'] + data['peigu']
* data['peigujia']) / (10 + data['peigu'] + data['songzhuangu'])
data['adj'] = (data['preclose'].shift(-1) /
data['close']).fillna(1)[::-1].cumprod()
data['open'] = data['open'] * data['adj']
data['high'] = data['high'] * data['adj']
data['low'] = data['low'] * data['adj']
data['close'] = data['close'] * data['adj']
data['preclose'] = data['preclose'] * data['adj']
data = data[data['if_trade']]
return data.drop(['fenhong', 'peigu', 'peigujia', 'songzhuangu', 'if_trade', 'category'], axis=1)[data['open'] != 0].assign(date=data['date'].apply(lambda x: str(x)[0:10]))[start_date:end_date]
elif if_fq in ['03', 'ddqfq']:
xdxr_data = QA_fetch_get_stock_xdxr(code)
info = xdxr_data[xdxr_data['category'] == 1]
bfq_data = data\
.assign(date=pd.to_datetime(data['datetime'].apply(lambda x: x[0:10])))\
.assign(code=str(code))\
.assign(date_stamp=data['datetime'].apply(lambda x: QA_util_date_stamp(str(x)[0:10])))\
.set_index('date', drop=False, inplace=False)\
.drop(['year', 'month', 'day', 'hour',
'minute', 'datetime'], axis=1)
bfq_data['if_trade']=True
data = pd.concat([bfq_data, info[['category']]
[bfq_data.index[0]:end_date]], axis=1)
data['date'] = data.index
data['if_trade'].fillna(value=False, inplace=True)
data = data.fillna(method='ffill')
data = pd.concat([data, info[['fenhong', 'peigu', 'peigujia',
'songzhuangu']][bfq_data.index[0]:end_date]], axis=1)
data = data.fillna(0)
data['preclose'] = (data['close'].shift(1) * 10 - data['fenhong'] + data['peigu']
* data['peigujia']) / (10 + data['peigu'] + data['songzhuangu'])
data['adj'] = (data['preclose'].shift(-1) /
data['close']).fillna(1)[::-1].cumprod()
data['open'] = data['open'] * data['adj']
data['high'] = data['high'] * data['adj']
data['low'] = data['low'] * data['adj']
data['close'] = data['close'] * data['adj']
data['preclose'] = data['preclose'] * data['adj']
data = data[data['if_trade']]
return data.drop(['fenhong', 'peigu', 'peigujia', 'songzhuangu', 'if_trade', 'category'], axis=1)[data['open'] != 0].assign(date=data['date'].apply(lambda x: str(x)[0:10]))[start_date:end_date]
elif if_fq in ['02', 'hfq']:
xdxr_data = QA_fetch_get_stock_xdxr(code)
info = xdxr_data[xdxr_data['category'] == 1]
bfq_data = data\
.assign(date=pd.to_datetime(data['datetime'].apply(lambda x: x[0:10])))\
.assign(code=str(code))\
.assign(date_stamp=data['datetime'].apply(lambda x: QA_util_date_stamp(str(x)[0:10])))\
.set_index('date', drop=False, inplace=False)\
.drop(['year', 'month', 'day', 'hour',
'minute', 'datetime'], axis=1)
bfq_data['if_trade']=True
data = pd.concat([bfq_data, info[['category']]
[bfq_data.index[0]:]], axis=1)
data['date'] = data.index
data['if_trade'].fillna(value=False, inplace=True)
data = data.fillna(method='ffill')
data = pd.concat([data, info[['fenhong', 'peigu', 'peigujia',
'songzhuangu']][bfq_data.index[0]:]], axis=1)
data = data.fillna(0)
data['preclose'] = (data['close'].shift(1) * 10 - data['fenhong'] + data['peigu']
* data['peigujia']) / (10 + data['peigu'] + data['songzhuangu'])
data['adj'] = (data['preclose'].shift(-1) /
data['close']).fillna(1).cumprod()
data['open'] = data['open'] / data['adj']
data['high'] = data['high'] / data['adj']
data['low'] = data['low'] / data['adj']
data['close'] = data['close'] / data['adj']
data['preclose'] = data['preclose'] / data['adj']
data = data[data['if_trade']]
return data.drop(['fenhong', 'peigu', 'peigujia', 'songzhuangu', 'if_trade', 'category'], axis=1)[data['open'] != 0].assign(date=data['date'].apply(lambda x: str(x)[0:10]))[start_date:end_date]
elif if_fq in ['04', 'ddhfq']:
xdxr_data = QA_fetch_get_stock_xdxr(code)
info = xdxr_data[xdxr_data['category'] == 1]
bfq_data = data\
.assign(date=pd.to_datetime(data['datetime'].apply(lambda x: x[0:10])))\
.assign(code=str(code))\
.assign(date_stamp=data['datetime'].apply(lambda x: QA_util_date_stamp(str(x)[0:10])))\
.set_index('date', drop=False, inplace=False)\
.drop(['year', 'month', 'day', 'hour',
'minute', 'datetime'], axis=1)
bfq_data['if_trade']=True
data = pd.concat([bfq_data, info[['category']]
[bfq_data.index[0]:end_date]], axis=1)
data['date'] = data.index
data['if_trade'].fillna(value=False, inplace=True)
data = data.fillna(method='ffill')
data = pd.concat([data, info[['fenhong', 'peigu', 'peigujia',
'songzhuangu']][bfq_data.index[0]:end_date]], axis=1)
data = data.fillna(0)
data['preclose'] = (data['close'].shift(1) * 10 - data['fenhong'] + data['peigu']
* data['peigujia']) / (10 + data['peigu'] + data['songzhuangu'])
data['adj'] = (data['preclose'].shift(-1) /
data['close']).fillna(1).cumprod()
data['open'] = data['open'] / data['adj']
data['high'] = data['high'] / data['adj']
data['low'] = data['low'] / data['adj']
data['close'] = data['close'] / data['adj']
data['preclose'] = data['preclose'] / data['adj']
data = data[data['if_trade']]
return data.drop(['fenhong', 'peigu', 'peigujia', 'songzhuangu', 'if_trade', 'category'], axis=1)[data['open'] != 0].assign(date=data['date'].apply(lambda x: str(x)[0:10]))[start_date:end_date]
def QA_fetch_get_stock_day(code, start_date, end_date, if_fq='00', level='day', ip=best_ip, port=7709):
api = TdxHq_API()
with api.connect(ip, port):
if level in ['day', 'd', 'D', 'DAY', 'Day']:
level = 9
elif level in ['w', 'W', 'Week', 'week']:
level = 5
elif level in ['month', 'M', 'm', 'Month']:
level = 6
elif level in ['Q', 'Quarter', 'q']:
level = 10
elif level in ['y', 'Y', 'year', 'Year']:
level = 11
data = pd.concat([api.to_df(api.get_security_bars(level, __select_market_code(
code), code, (9 - i) * 800, 800)) for i in range(10)], axis=0)
if if_fq in ['00', 'bfq']:
data = data.assign(date=data['datetime'].apply(lambda x: str(x[0:10]))).assign(code=str(code))\
.assign(date_stamp=data['datetime'].apply(lambda x: QA_util_date_stamp(str(x)[0:10])))\
.set_index('date', drop=False, inplace=False)\
.drop(['year', 'month', 'day', 'hour',
'minute', 'datetime'], axis=1)[start_date:end_date]
return data.assign(date=data['date'].apply(lambda x: str(x)[0:10]))
elif if_fq in ['01', 'qfq']:
xdxr_data = QA_fetch_get_stock_xdxr(code)
info = xdxr_data[xdxr_data['category'] == 1]
bfq_data = data\
.assign(date=pd.to_datetime(data['datetime'].apply(lambda x: x[0:10])))\
.assign(code=str(code))\
.assign(date_stamp=data['datetime'].apply(lambda x: QA_util_date_stamp(str(x)[0:10])))\
.set_index('date', drop=False, inplace=False)\
.drop(['year', 'month', 'day', 'hour',
'minute', 'datetime'], axis=1)
bfq_data['if_trade'] = True
data = pd.concat([bfq_data, info[['category']]
[bfq_data.index[0]:]], axis=1)
data['date'] = data.index
data['if_trade'].fillna(value=False, inplace=True)
data = data.fillna(method='ffill')
data = pd.concat([data, info[['fenhong', 'peigu', 'peigujia',
'songzhuangu']][bfq_data.index[0]:]], axis=1)
data = data.fillna(0)
data['preclose'] = (data['close'].shift(1) * 10 - data['fenhong'] + data['peigu']
* data['peigujia']) / (10 + data['peigu'] + data['songzhuangu'])
data['adj'] = (data['preclose'].shift(-1) /
data['close']).fillna(1)[::-1].cumprod()
data['open'] = data['open'] * data['adj']
data['high'] = data['high'] * data['adj']
data['low'] = data['low'] * data['adj']
data['close'] = data['close'] * data['adj']
data['preclose'] = data['preclose'] * data['adj']
data = data[data['if_trade']]
return data.drop(['fenhong', 'peigu', 'peigujia', 'songzhuangu', 'if_trade', 'category'], axis=1)[data['open'] != 0].assign(date=data['date'].apply(lambda x: str(x)[0:10]))[start_date:end_date]
elif if_fq in ['03', 'ddqfq']:
xdxr_data = QA_fetch_get_stock_xdxr(code)
info = xdxr_data[xdxr_data['category'] == 1]
bfq_data = data\
.assign(date=pd.to_datetime(data['datetime'].apply(lambda x: x[0:10])))\
.assign(code=str(code))\
.assign(date_stamp=data['datetime'].apply(lambda x: QA_util_date_stamp(str(x)[0:10])))\
.set_index('date', drop=False, inplace=False)\
.drop(['year', 'month', 'day', 'hour',
'minute', 'datetime'], axis=1)
bfq_data['if_trade'] = True
data = pd.concat([bfq_data, info[['category']]
[bfq_data.index[0]:end_date]], axis=1)
data['date'] = data.index
data['if_trade'].fillna(value=False, inplace=True)
data = data.fillna(method='ffill')
data = pd.concat([data, info[['fenhong', 'peigu', 'peigujia',
'songzhuangu']][bfq_data.index[0]:end_date]], axis=1)
data = data.fillna(0)
data['preclose'] = (data['close'].shift(1) * 10 - data['fenhong'] + data['peigu']
* data['peigujia']) / (10 + data['peigu'] + data['songzhuangu'])
data['adj'] = (data['preclose'].shift(-1) /
data['close']).fillna(1)[::-1].cumprod()
data['open'] = data['open'] * data['adj']
data['high'] = data['high'] * data['adj']
data['low'] = data['low'] * data['adj']
data['close'] = data['close'] * data['adj']
data['preclose'] = data['preclose'] * data['adj']
data = data[data['if_trade']]
return data.drop(['fenhong', 'peigu', 'peigujia', 'songzhuangu', 'if_trade', 'category'], axis=1)[data['open'] != 0].assign(date=data['date'].apply(lambda x: str(x)[0:10]))[start_date:end_date]
elif if_fq in ['02', 'hfq']:
xdxr_data = QA_fetch_get_stock_xdxr(code)
info = xdxr_data[xdxr_data['category'] == 1]
bfq_data = data\
.assign(date=pd.to_datetime(data['datetime'].apply(lambda x: x[0:10])))\
.assign(code=str(code))\
.assign(date_stamp=data['datetime'].apply(lambda x: QA_util_date_stamp(str(x)[0:10])))\
.set_index('date', drop=False, inplace=False)\
.drop(['year', 'month', 'day', 'hour',
'minute', 'datetime'], axis=1)
bfq_data['if_trade'] = True
data = pd.concat([bfq_data, info[['category']]
[bfq_data.index[0]:]], axis=1)
data['date'] = data.index
data['if_trade'].fillna(value=False, inplace=True)
data = data.fillna(method='ffill')
data = pd.concat([data, info[['fenhong', 'peigu', 'peigujia',
'songzhuangu']][bfq_data.index[0]:]], axis=1)
data = data.fillna(0)
data['preclose'] = (data['close'].shift(1) * 10 - data['fenhong'] + data['peigu']
* data['peigujia']) / (10 + data['peigu'] + data['songzhuangu'])
data['adj'] = (data['preclose'].shift(-1) /
data['close']).fillna(1).cumprod()
data['open'] = data['open'] / data['adj']
data['high'] = data['high'] / data['adj']
data['low'] = data['low'] / data['adj']
data['close'] = data['close'] / data['adj']
data['preclose'] = data['preclose'] / data['adj']
data = data[data['if_trade']]
return data.drop(['fenhong', 'peigu', 'peigujia', 'songzhuangu', 'if_trade', 'category'], axis=1)[data['open'] != 0].assign(date=data['date'].apply(lambda x: str(x)[0:10]))[start_date:end_date]
elif if_fq in ['04', 'ddhfq']:
xdxr_data = QA_fetch_get_stock_xdxr(code)
info = xdxr_data[xdxr_data['category'] == 1]
bfq_data = data\
.assign(date=pd.to_datetime(data['datetime'].apply(lambda x: x[0:10])))\
.assign(code=str(code))\
.assign(date_stamp=data['datetime'].apply(lambda x: QA_util_date_stamp(str(x)[0:10])))\
.set_index('date', drop=False, inplace=False)\
.drop(['year', 'month', 'day', 'hour',
'minute', 'datetime'], axis=1)
bfq_data['if_trade'] = True
data = pd.concat([bfq_data, info[['category']]
[bfq_data.index[0]:end_date]], axis=1)
data['date'] = data.index
data['if_trade'].fillna(value=False, inplace=True)
data = data.fillna(method='ffill')
data = pd.concat([data, info[['fenhong', 'peigu', 'peigujia',
'songzhuangu']][bfq_data.index[0]:end_date]], axis=1)
data = data.fillna(0)
data['preclose'] = (data['close'].shift(1) * 10 - data['fenhong'] + data['peigu']
* data['peigujia']) / (10 + data['peigu'] + data['songzhuangu'])
data['adj'] = (data['preclose'].shift(-1) /
data['close']).fillna(1).cumprod()
data['open'] = data['open'] / data['adj']
data['high'] = data['high'] / data['adj']
data['low'] = data['low'] / data['adj']
data['close'] = data['close'] / data['adj']
data['preclose'] = data['preclose'] / data['adj']
data = data[data['if_trade']]
return data.drop(['fenhong', 'peigu', 'peigujia', 'songzhuangu', 'if_trade', 'category'], axis=1)[data['open'] != 0].assign(date=data['date'].apply(lambda x: str(x)[0:10]))[start_date:end_date]
def QA_fetch_get_stock_day(code, start_date, end_date, if_fq='00', frequence='day', ip=None, port=None):
"""获取日线及以上级别的数据
Arguments:
code {str:6} -- code 是一个单独的code 6位长度的str
start_date {str:10} -- 10位长度的日期 比如'2017-01-01'
end_date {str:10} -- 10位长度的日期 比如'2018-01-01'
Keyword Arguments:
if_fq {str} -- '00'/'bfq' -- 不复权 '01'/'qfq' -- 前复权 '02'/'hfq' -- 后复权 '03'/'ddqfq' -- 定点前复权 '04'/'ddhfq' --定点后复权
frequency {str} -- day/week/month/quarter/year 也可以是简写 D/W/M/Q/Y
ip {str} -- [description] (default: None) ip可以通过select_best_ip()函数重新获取
port {int} -- [description] (default: {None})
Returns:
pd.DataFrame/None -- 返回的是dataframe,如果出错比如只获取了一天,而当天停牌,返回None
Exception:
如果出现网络问题/服务器拒绝, 会出现socket:time out 尝试再次获取/更换ip即可, 本函数不做处理
"""
global best_ip
if ip is None and port is None and best_ip['stock']['ip'] is None and best_ip['stock']['port'] is None:
best_ip = select_best_ip()
ip = best_ip['stock']['ip']
port = best_ip['stock']['port']
elif ip is None and port is None and best_ip['stock']['ip'] is not None and best_ip['stock']['port'] is not None:
ip = best_ip['stock']['ip']
port = best_ip['stock']['port']
else:
pass
api = TdxHq_API()
with api.connect(ip, port, time_out=0.7):
if frequence in ['day', 'd', 'D', 'DAY', 'Day']:
frequence = 9
elif frequence in ['w', 'W', 'Week', 'week']:
frequence = 5
elif frequence in ['month', 'M', 'm', 'Month']:
frequence = 6
elif frequence in ['quarter', 'Q', 'Quarter', 'q']:
frequence = 10
elif frequence in ['y', 'Y', 'year', 'Year']:
frequence = 11
start_date = str(start_date)[0:10]
today_ = datetime.date.today()
lens = QA_util_get_trade_gap(start_date, today_)
data = pd.concat([api.to_df(api.get_security_bars(frequence, _select_market_code(
code), code, (int(lens / 800) - i) * 800, 800)) for i in range(int(lens / 800) + 1)], axis=0)
# 这里的问题是: 如果只取了一天的股票,而当天停牌, 那么就直接返回None了
if len(data) < 1:
return None
data = data[data['open'] != 0]
if if_fq in ['00', 'bfq']:
data = data.assign(date=data['datetime'].apply(lambda x: str(x[0:10]))).assign(code=str(code))\
.assign(date_stamp=data['datetime'].apply(lambda x: QA_util_date_stamp(str(x)[0:10]))).set_index('date', drop=False, inplace=False)
return data.drop(['year', 'month', 'day', 'hour', 'minute', 'datetime'], axis=1)[start_date:end_date].assign(date=data['date'].apply(lambda x: str(x)[0:10]))
elif if_fq in ['01', 'qfq']:
xdxr_data = QA_fetch_get_stock_xdxr(code)
bfq_data = data.assign(date=pd.to_datetime(data['datetime'].apply(lambda x: str(x[0:10])))).assign(code=str(code))\
.assign(date_stamp=data['datetime'].apply(lambda x: QA_util_date_stamp(str(x)[0:10]))).set_index('date', drop=False, inplace=False)
bfq_data = bfq_data.drop(
['year', 'month', 'day', 'hour', 'minute', 'datetime'], axis=1)
#
if xdxr_data is not None:
info = xdxr_data[xdxr_data['category'] == 1]
bfq_data['if_trade'] = True
data = pd.concat([bfq_data, info[['category']]
[bfq_data.index[0]:]], axis=1)
data['date'] = data.index
data['if_trade'].fillna(value=False, inplace=True)
data = data.fillna(method='ffill')
data = pd.concat([data, info[['fenhong', 'peigu', 'peigujia',
'songzhuangu']][bfq_data.index[0]:]], axis=1)
data = data.fillna(0)
data['preclose'] = (data['close'].shift(1) * 10 - data['fenhong'] + data['peigu']
* data['peigujia']) / (10 + data['peigu'] + data['songzhuangu'])
data['adj'] = (data['preclose'].shift(-1) /
data['close']).fillna(1)[::-1].cumprod()
data['open'] = data['open'] * data['adj']
data['high'] = data['high'] * data['adj']
data['low'] = data['low'] * data['adj']
data['close'] = data['close'] * data['adj']
data['preclose'] = data['preclose'] * data['adj']
data = data[data['if_trade']]
return data.drop(['fenhong', 'peigu', 'peigujia', 'songzhuangu', 'if_trade', 'category'], axis=1)[data['open'] != 0].assign(date=data['date'].apply(lambda x: str(x)[0:10]))[start_date:end_date]
else:
bfq_data['preclose'] = bfq_data['close'].shift(1)
bfq_data['adj'] = 1
return bfq_data[start_date:end_date]
elif if_fq in ['03', 'ddqfq']:
xdxr_data = QA_fetch_get_stock_xdxr(code)
info = xdxr_data[xdxr_data['category'] == 1]
bfq_data = data\
.assign(date=pd.to_datetime(data['datetime'].apply(lambda x: x[0:10])))\
.assign(code=str(code))\
.assign(date_stamp=data['datetime'].apply(lambda x: QA_util_date_stamp(str(x)[0:10])))\
.set_index('date', drop=False, inplace=False)\
.drop(['year', 'month', 'day', 'hour',
'minute', 'datetime'], axis=1)
bfq_data['if_trade'] = True
data = pd.concat([bfq_data, info[['category']]
[bfq_data.index[0]:end_date]], axis=1)
data['date'] = data.index
data['if_trade'].fillna(value=False, inplace=True)
data = data.fillna(method='ffill')
data = pd.concat([data, info[['fenhong', 'peigu', 'peigujia',
'songzhuangu']][bfq_data.index[0]:end_date]], axis=1)
data = data.fillna(0)
data['preclose'] = (data['close'].shift(1) * 10 - data['fenhong'] + data['peigu']
* data['peigujia']) / (10 + data['peigu'] + data['songzhuangu'])
data['adj'] = (data['preclose'].shift(-1) /
data['close']).fillna(1)[::-1].cumprod()
data['open'] = data['open'] * data['adj']
data['high'] = data['high'] * data['adj']
data['low'] = data['low'] * data['adj']
data['close'] = data['close'] * data['adj']
data['preclose'] = data['preclose'] * data['adj']
data = data[data['if_trade']]
return data.drop(['fenhong', 'peigu', 'peigujia', 'songzhuangu', 'if_trade', 'category'], axis=1)[data['open'] != 0].assign(date=data['date'].apply(lambda x: str(x)[0:10]))[start_date:end_date]
elif if_fq in ['02', 'hfq']:
xdxr_data = QA_fetch_get_stock_xdxr(code)
info = xdxr_data[xdxr_data['category'] == 1]
bfq_data = data\
.assign(date=pd.to_datetime(data['datetime'].apply(lambda x: x[0:10])))\
.assign(code=str(code))\
.assign(date_stamp=data['datetime'].apply(lambda x: QA_util_date_stamp(str(x)[0:10])))\
.set_index('date', drop=False, inplace=False)\
.drop(['year', 'month', 'day', 'hour',
'minute', 'datetime'], axis=1)
bfq_data['if_trade'] = True
data = pd.concat([bfq_data, info[['category']]
[bfq_data.index[0]:]], axis=1)
data['date'] = data.index
data['if_trade'].fillna(value=False, inplace=True)
data = data.fillna(method='ffill')
data = pd.concat([data, info[['fenhong', 'peigu', 'peigujia',
'songzhuangu']][bfq_data.index[0]:]], axis=1)
data = data.fillna(0)
data['preclose'] = (data['close'].shift(1) * 10 - data['fenhong'] + data['peigu']
* data['peigujia']) / (10 + data['peigu'] + data['songzhuangu'])
data['adj'] = (data['preclose'].shift(-1) /
data['close']).fillna(1).cumprod()
data['open'] = data['open'] / data['adj']
data['high'] = data['high'] / data['adj']
data['low'] = data['low'] / data['adj']
data['close'] = data['close'] / data['adj']
data['preclose'] = data['preclose'] / data['adj']
data = data[data['if_trade']]
return data.drop(['fenhong', 'peigu', 'peigujia', 'songzhuangu', 'if_trade', 'category'], axis=1)[data['open'] != 0].assign(date=data['date'].apply(lambda x: str(x)[0:10]))[start_date:end_date]
elif if_fq in ['04', 'ddhfq']:
xdxr_data = QA_fetch_get_stock_xdxr(code)
info = xdxr_data[xdxr_data['category'] == 1]
bfq_data = data\
.assign(date=pd.to_datetime(data['datetime'].apply(lambda x: x[0:10])))\
.assign(code=str(code))\
.assign(date_stamp=data['datetime'].apply(lambda x: QA_util_date_stamp(str(x)[0:10])))\
.set_index('date', drop=False, inplace=False)\
.drop(['year', 'month', 'day', 'hour',
'minute', 'datetime'], axis=1)
bfq_data['if_trade'] = True
data = pd.concat([bfq_data, info[['category']]
[bfq_data.index[0]:end_date]], axis=1)
data['date'] = data.index
data['if_trade'].fillna(value=False, inplace=True)
data = data.fillna(method='ffill')
data = pd.concat([data, info[['fenhong', 'peigu', 'peigujia',
'songzhuangu']][bfq_data.index[0]:end_date]], axis=1)
data = data.fillna(0)
data['preclose'] = (data['close'].shift(1) * 10 - data['fenhong'] + data['peigu']
* data['peigujia']) / (10 + data['peigu'] + data['songzhuangu'])
data['adj'] = (data['preclose'].shift(-1) /
data['close']).fillna(1).cumprod()
data['open'] = data['open'] / data['adj']
data['high'] = data['high'] / data['adj']
data['low'] = data['low'] / data['adj']
data['close'] = data['close'] / data['adj']
data['preclose'] = data['preclose'] / data['adj']
data = data[data['if_trade']]
return data.drop(['fenhong', 'peigu', 'peigujia', 'songzhuangu', 'if_trade', 'category'], axis=1)[data['open'] != 0].assign(date=data['date'].apply(lambda x: str(x)[0:10]))[start_date:end_date]
def test_all_functions(multithread, heartbeat, auto_retry, raise_exception):
api = TdxHq_API(multithread=multithread, heartbeat=heartbeat,
auto_retry=auto_retry, raise_exception=raise_exception)
with api.connect(time_out=30):
log.info("获取股票行情")
stocks = api.get_security_quotes([(0, "000001"), (1, "600300")])
assert stocks is not None
assert type(stocks) is list
# 方法2
stocks = api.get_security_quotes(0, "000001")
assert stocks is not None
assert type(stocks) is list
# 方法3
stocks = api.get_security_quotes((0, "000001"))
assert stocks is not None
assert type(stocks) is list
log.info("获取k线")
data = api.get_security_bars(9, 0, '000001', 4, 3)
assert data is not None
assert type(data) is list
assert len(data) == 3
log.info("获取 深市 股票数量")
assert api.get_security_count(0) > 0
log.info("获取股票列表")
stocks = api.get_security_list(1, 0)
assert stocks is not None
assert type(stocks) is list
assert len(stocks) > 0
log.info("获取指数k线")
data = api.get_index_bars(9, 1, '000001', 1, 2)
assert data is not None
assert type(data) is list
assert len(data) == 2
log.info("查询分时行情")
data = api.get_minute_time_data(TDXParams.MARKET_SH, '600300')
assert data is not None
log.info("查询历史分时行情")
data = api.get_history_minute_time_data(
TDXParams.MARKET_SH, '600300', 20161209)
assert data is not None
assert type(data) is list
assert len(data) > 0
log.info("查询分时成交")
data = api.get_transaction_data(TDXParams.MARKET_SZ, '000001', 0, 30)
assert data is not None
assert type(data) is list
log.info("查询历史分时成交")
data = api.get_history_transaction_data(
TDXParams.MARKET_SZ, '000001', 0, 10, 20170209)
assert data is not None
assert type(data) is list
assert len(data) == 10
log.info("查询公司信息目录")
data = api.get_company_info_category(TDXParams.MARKET_SZ, '000001')
assert data is not None
assert type(data) is list
assert len(data) > 0
start = data[0]['start']
length = data[0]['length']
log.info("读取公司信息-最新提示")
data = api.get_company_info_content(
0, '000001', '000001.txt', start, length)
assert data is not None
assert len(data) > 0
log.info("读取除权除息信息")
data = api.get_xdxr_info(1, '600300')
assert data is not None
assert type(data) is list
assert len(data) > 0
log.info("读取财务信息")
data = api.get_finance_info(0, '000001')
assert data is not None
assert type(data) is OrderedDict
assert len(data) > 0
log.info("日线级别k线获取函数")
data = api.get_k_data('000001', '2017-07-01', '2017-07-10')
assert type(data) is pd.DataFrame
assert len(data) == 6
log.info("获取板块信息")
data = api.get_and_parse_block_info(TDXParams.BLOCK_FG)
assert data is not None
assert type(data) is list
assert len(data) > 0
class TBStockData:
__serverList = []
_bestIP = []
__bestIPFile = ''
__tdx = None
_lastBaseHistList = pd.DataFrame()
_xdxrData = None
def __init__(self, autoIP = False):
self.__serverList = hq_hosts
self.__bestIPFile = os.path.dirname(os.path.realpath(__file__)) + '/best.ip'
if autoIP:
self.getBestIP()
else:
if os.path.exists(self.__bestIPFile):
with open(self.__bestIPFile, 'r') as f:
data = f.read()
self._bestIP = json.loads(data)
def ping(self, ip, port):
api = TdxHq_API()
time1 = datetime.datetime.now()
try:
with api.connect(ip, int(port)):
if len(api.get_security_list(0, 1)) > 800:
return datetime.datetime.now() - time1
else:
return datetime.timedelta(9, 9, 0)
except:
return datetime.timedelta(9, 9, 0)
def getBestIP(self):
pingTimeList = [self.ping(x[1], x[2]) for x in self.__serverList]
self._bestIP = self.__serverList[pingTimeList.index(min(pingTimeList))]
with open(self.__bestIPFile, 'w') as f:
f.write(json.dumps(self._bestIP))
def showAllIP(self):
for item in self.__serverList:
print item[0],'\t', item[1], '\t', item[2]
def _connect(self):
if self.__tdx is None:
if not self._bestIP:
self.getBestIP()
#self.__tdx = TdxHq_API(heartbeat=True, auto_retry=True)
self.__tdx = TdxHq_API(auto_retry=True)
self.__tdx.connect(self._bestIP[1], int(self._bestIP[2]))
#计算量比
def _setVolRaito(self, row):
date = row.name
histList = self._lastBaseHistList[:date]
if len(histList) < 6:
return np.nan
return round((histList['vol'].values[-1] / 240) / (histList[-6:-1]['vol'].sum() / 1200), 3)
#计算各种指标
def getData(self, df = pd.DataFrame(), indexs=['turnover', 'vol', 'ma', 'macd', 'kdj', 'cci', 'bbi', 'sar', 'trix']):
indexs = [x.lower() for x in indexs]
histList = pd.DataFrame()
if not df.empty:
histList = df.copy()
elif not self._lastBaseHistList.empty:
histList = self._lastBaseHistList.copy()
if histList.empty:
return None
dayKStatus = False
try:
if int(time.mktime(time.strptime(str(histList.index[-1]), "%Y-%m-%d %X"))) - int(time.mktime(time.strptime(str(histList.index[-2]), "%Y-%m-%d %X"))) > 43200:
#日线以上行情
dayKStatus = True
except:
dayKStatus = True
#计算涨幅
histList['p_change'] = histList['close'].pct_change().round(5) * 100
#量比
histList['vol_ratio'] = histList.apply(self._setVolRaito, axis=1)
#振幅
histList['amp'] = ((histList['high'] - histList['low']) / histList.shift()['close'] * 100).round(3)
#计算换手率
if self._xdxrData is None:
xdxrData = self.getXdxr(str(histList['code'].values[0]))
else:
xdxrData = self._xdxrData
info = xdxrData[xdxrData['liquidity_after'] > 0][['liquidity_after', 'shares_after']]
if dayKStatus:
startDate = str(histList.index[0])[0:10]
endDate = str(histList.index[-1])[0:10]
info1 = info[info.index <= startDate][-1:]
info = info1.append(info[info.index >= startDate]).drop_duplicates()
info = info.reindex(pd.date_range(info1.index[-1], endDate))
info = info.resample('1D').last().fillna(method='pad')[startDate:endDate]
#info['date'] = info.index
#info['date'] = info['date'].dt.strftime('%Y-%m-%d')
#info = info.set_index('date')
circulate = info['liquidity_after'] * 10000
capital = info['shares_after'] * 10000
else:
circulate = info['liquidity_after'].values[-1] * 10000
capital = info['shares_after'].values[-1] * 10000
#histList['circulate'] = (circulate / 10000 / 10000).round(4)
if 'turnover' in indexs and dayKStatus:
histList['turnover'] = (histList['vol'] * 100 / circulate).round(5) * 100
histList['turnover5'] = talib.MA(histList['turnover'].values, timeperiod=5).round(3)
#stockstats转换,主要是用来计算KDJ等相关指标
#用talib计算KDJ时会与现有软件偏差大
ss = StockDataFrame.retype(histList[['high','low','open','close']])
#MACD计算
if 'macd' in indexs:
difList, deaList, macdList = talib.MACD(histList['close'].values, fastperiod=12, slowperiod=26, signalperiod=9)
macdList = macdList * 2
histList['macd_dif'] = difList.round(3)
histList['macd_dea'] = deaList.round(3)
histList['macd_value'] = macdList.round(3)
histList['macd_value_ma'] = 0
try:
histList['macd_value_ma'] = talib.MA(histList['macd_value'].values, timeperiod=5).round(3)
except:
pass
histList['macd_cross_status'] = 0
macdPosList = histList['macd_dif'] > histList['macd_dea']
histList.loc[macdPosList[(macdPosList == True) & (macdPosList.shift() == False)].index, 'macd_cross_status'] = 1
histList.loc[macdPosList[(macdPosList == False) & (macdPosList.shift() == True)].index, 'macd_cross_status'] = -1
#histList[['macd_cross_status']] = histList[['macd_cross_status']].fillna(method='pad')
#KDJ计算
if 'kdj' in indexs:
histList['kdj_k'] = ss['kdjk'].round(3)
histList['kdj_d'] = ss['kdjd'].round(3)
histList['kdj_j'] = ss['kdjj'].round(3)
histList['kdj_cross_status'] = 0
kdjPosList = histList['kdj_k'] >= histList['kdj_d']
histList.loc[kdjPosList[(kdjPosList == True) & (kdjPosList.shift() == False)].index, 'kdj_cross_status'] = 1
histList.loc[kdjPosList[(kdjPosList == False) & (kdjPosList.shift() == True)].index, 'kdj_cross_status'] = -1
#histList[['kdj_cross_status']] = histList[['kdj_cross_status']].fillna(method='pad')
#CCI计算
if 'cci' in indexs:
histList['cci'] = ss['cci'].round(3)
#ma相关计算
if 'ma' in indexs:
histList['ma5'] = talib.MA(histList['close'].values, timeperiod=5).round(3)
histList['ma10'] = talib.MA(histList['close'].values, timeperiod=10).round(3)
histList['ma20'] = talib.MA(histList['close'].values, timeperiod=20).round(3)
histList['ma30'] = talib.MA(histList['close'].values, timeperiod=30).round(3)
histList['ma60'] = talib.MA(histList['close'].values, timeperiod=60).round(3)
histList['ma240'] = talib.MA(histList['close'].values, timeperiod=240).round(3)
histList[['ma5', 'ma10', 'ma20', 'ma30', 'ma60', 'ma240']] = histList[['ma5', 'ma10', 'ma20', 'ma30', 'ma60', 'ma240']].fillna(0)
#成交量计算
if 'vol' in indexs:
histList['vol5'] = talib.MA(histList['vol'].values, timeperiod=5).round(3)
histList['vol10'] = talib.MA(histList['vol'].values, timeperiod=10).round(3)
histList['vol20'] = talib.MA(histList['vol'].values, timeperiod=20).round(3)
histList['vol_zoom'] = (histList['vol'] / histList['vol5'] * 1.0).round(3)
histList['vol5_vol10_cross_status'] = 0
volumePosList = histList['vol5'] >= histList['vol10']
histList.loc[volumePosList[(volumePosList == True) & (volumePosList.shift() == False)].index, 'vol5_vol10_cross_status'] = 1
histList.loc[volumePosList[(volumePosList == False) & (volumePosList.shift() == True)].index, 'vol5_vol10_cross_status'] = -1
del volumePosList
histList['vol5_vol20_cross_status'] = 0
volumePosList = histList['vol5'] >= histList['vol20']
histList.loc[volumePosList[(volumePosList == True) & (volumePosList.shift() == False)].index, 'vol5_vol20_cross_status'] = 1
histList.loc[volumePosList[(volumePosList == False) & (volumePosList.shift() == True)].index, 'vol5_vol20_cross_status'] = -1
del volumePosList
histList['vol10_vol20_cross_status'] = 0
volumePosList = histList['vol10'] >= histList['vol20']
histList.loc[volumePosList[(volumePosList == True) & (volumePosList.shift() == False)].index, 'vol10_vol20_cross_status'] = 1
histList.loc[volumePosList[(volumePosList == False) & (volumePosList.shift() == True)].index, 'vol10_vol20_cross_status'] = -1
#histList[['vol5_vol10_cross_status', 'vol5_vol20_cross_status', 'vol10_vol20_cross_status']] = histList[['vol5_vol10_cross_status', 'vol5_vol20_cross_status', 'vol10_vol20_cross_status']].fillna(method='pad')
#bbi计算
if 'bbi' in indexs:
ma3 = talib.MA(histList['close'].values, timeperiod=3)
ma6 = talib.MA(histList['close'].values, timeperiod=6)
ma12 = talib.MA(histList['close'].values, timeperiod=12)
ma24 = talib.MA(histList['close'].values, timeperiod=24)
histList['bbi'] = (ma3 + ma6 + ma12 + ma24) / 4
histList['bbi'] = histList['bbi'].round(3)
#SAR计算
if 'sar' in indexs:
sarList = talib.SAR(histList['high'].values, histList['low'].values, acceleration=0.04, maximum=0.2)
histList['sar'] = sarList.round(3)
histList['sar_cross_status'] = 0
sarPosList = histList['close'] >= histList['sar']
histList.loc[sarPosList[(sarPosList == True) & (sarPosList.shift() == False)].index, 'sar_cross_status'] = 1
histList.loc[sarPosList[(sarPosList == False) & (sarPosList.shift() == True)].index, 'sar_cross_status'] = -1
#计算TRIX
if 'trix' in indexs:
histList['trix'] = np.nan
histList['trma'] = np.nan
histList['trix_diff'] = np.nan
try:
trix = talib.TRIX(histList['close'].values, 12)
trma = talib.MA(trix, timeperiod=20)
histList['trix'] = trix.round(3)
histList['trma'] = trma.round(3)
histList['trix_diff'] = histList['trix'] - histList['trma']
histList['trix_cross_status'] = 0
trixPosList = histList['trix'] >= histList['trma']
histList.loc[trixPosList[(trixPosList == True) & (trixPosList.shift() == False)].index, 'trix_cross_status'] = 1
histList.loc[trixPosList[(trixPosList == False) & (trixPosList.shift() == True)].index, 'trix_cross_status'] = -1
#histList[['trix_cross_status']] = histList[['trix_cross_status']].fillna(method='pad')
except:
pass
if 'cyc' in indexs:
avePrice = histList['amount'] / (histList['vol'] * 100)
histList['cyc5'] = talib.MA(avePrice.values, timeperiod=5).round(3)
histList['cyc13'] = talib.MA(avePrice.values, timeperiod=13).round(3)
histList['cyc34'] = talib.MA(avePrice.values, timeperiod=34).round(3)
#histList['cycx'] = talib.EMA(histList['close'].values, timeperiod=histList['vol'].values * 100 / circulate).round(3)
histList['cyc5_cyc13_cross_status'] = 0
volumePosList = histList['cyc5'] >= histList['cyc13']
histList.loc[volumePosList[(volumePosList == True) & (volumePosList.shift() == False)].index, 'cyc5_cyc13_cross_status'] = 1
histList.loc[volumePosList[(volumePosList == False) & (volumePosList.shift() == True)].index, 'cyc5_cyc13_cross_status'] = -1
del volumePosList
histList['cyc13_cyc34_cross_status'] = 0
volumePosList = histList['cyc13'] >= histList['cyc34']
histList.loc[volumePosList[(volumePosList == True) & (volumePosList.shift() == False)].index, 'cyc13_cyc34_cross_status'] = 1
histList.loc[volumePosList[(volumePosList == False) & (volumePosList.shift() == True)].index, 'cyc13_cyc34_cross_status'] = -1
del volumePosList
if 'boll' in indexs:
up, mid, low = talib.BBANDS(
histList['close'].values,
timeperiod=20,
# number of non-biased standard deviations from the mean
nbdevup=2,
nbdevdn=2,
# Moving average type: simple moving average here
matype=0)
histList['boll_up'] = up.round(3)
histList['boll_mid'] = mid.round(3)
histList['boll_low'] = low.round(3)
return histList
#整理开始,结束时间,并计算相差天数
def _getDate(self, start, end):
if not end:
end = time.strftime('%Y-%m-%d',time.localtime())
if not start:
t = int(time.mktime(time.strptime(str(end), '%Y-%m-%d'))) - 86400 * 800
start = str(time.strftime('%Y-%m-%d',time.localtime(t)))
startTimestamp = int(time.mktime(time.strptime(str(start), '%Y-%m-%d')))
endTimestamp = int(time.mktime(time.strptime(str(end), '%Y-%m-%d')))
diffDayNum = int((time.time() - startTimestamp) / 86400)
if diffDayNum <= 0:
diffDayNum = 1
return start, end, diffDayNum
#得到市场代码
def getMarketCode(self, code):
code = str(code)
if code[0] in ['5', '6', '9'] or code[:3] in ["009", "126", "110", "201", "202", "203", "204"]:
return 1
return 0
#时间整理
def _dateStamp(self, date):
datestr = str(date)[0:10]
date = time.mktime(time.strptime(datestr, '%Y-%m-%d'))
return date
#整理时间
def _timeStamp(self, _time):
if len(str(_time)) == 10:
# yyyy-mm-dd格式
return time.mktime(time.strptime(_time, '%Y-%m-%d'))
elif len(str(_time)) == 16:
# yyyy-mm-dd hh:mm格式
return time.mktime(time.strptime(_time, '%Y-%m-%d %H:%M'))
else:
timestr = str(_time)[0:19]
return time.mktime(time.strptime(timestr, '%Y-%m-%d %H:%M:%S'))
#得到除权信息
def getXdxr(self, code):
self._connect()
category = {
'1': '除权除息', '2': '送配股上市', '3': '非流通股上市', '4': '未知股本变动', '5': '股本变化',
'6': '增发新股', '7': '股份回购', '8': '增发新股上市', '9': '转配股上市', '10': '可转债上市',
'11': '扩缩股', '12': '非流通股缩股', '13': '送认购权证', '14': '送认沽权证'}
data = self.__tdx.to_df(self.__tdx.get_xdxr_info(self.getMarketCode(code), code))
if len(data) >= 1:
data = data\
.assign(date=pd.to_datetime(data[['year', 'month', 'day']], format='%Y-%m-%d'))\
.drop(['year', 'month', 'day'], axis=1)\
.assign(category_meaning=data['category'].apply(lambda x: category[str(x)]))\
.assign(code=str(code))\
.rename(index=str, columns={'panhouliutong': 'liquidity_after',
'panqianliutong': 'liquidity_before', 'houzongguben': 'shares_after',
'qianzongguben': 'shares_before'})\
.set_index('date', drop=False, inplace=False)
xdxrData = data.assign(date=data['date'].apply(lambda x: str(x)[0:10]))
#xdxrData = xdxrData.set_index('date')
self._xdxrData = xdxrData
return xdxrData
else:
return None
#得到股本
def getGuben(self, code):
self._connect()
if self._xdxrData is None:
xdxrData = self.getXdxr(code)
else:
xdxrData = self._xdxrData
info = xdxrData[xdxrData['liquidity_after'] > 0][['liquidity_after', 'shares_after']]
circulate = info['liquidity_after'].values[-1] * 10000
capital = info['shares_after'].values[-1] * 10000
return capital,circulate
#按天得到标准数据
'''
ktype = D(天)/W(周)/M(月)/Q(季)/Y(年)
autype = bfq(不复权)/hfq(后复权)/qfq(前复权)
'''
def getDays(self, code, ktype = 'D', start = '', end = '', autype = 'qfq', indexs = ['turnover', 'vol', 'ma', 'macd', 'kdj', 'cci', 'bbi', 'sar', 'trix']):
startDate, endDate, diffDayNum = self._getDate(start, end)
self._connect()
ktypeCode = 9
if ktype.lower() == 'd':
ktypeCode = 9
elif ktype.lower() == 'w':
ktypeCode = 5
elif ktype.lower() == 'm':
ktypeCode = 6
elif ktype.lower() == 'q':
ktypeCode = 10
elif ktype.lower() == 'y':
ktypeCode = 11
histList = pd.concat([self.__tdx.to_df(self.__tdx.get_security_bars(ktypeCode, self.getMarketCode(code), code, (int(diffDayNum / 800) - i) * 800, 800)) for i in range(int(diffDayNum / 800) + 1)], axis=0)
if histList.empty:
return None
histList = histList[histList['open'] != 0]
histList = histList[histList['vol'] > 1]
if not autype or autype == 'bfq':
histList = histList.assign(date=histList['datetime'].apply(lambda x: str(x[0:10]))).assign(code=str(code))\
.assign(date_stamp=histList['datetime'].apply(lambda x: self._dateStamp(str(x)[0:10])))
histList = histList.drop(['year', 'month', 'day', 'hour', 'minute', 'datetime', 'date_stamp'], axis=1)
histList = histList.set_index('date')
histList = histList[startDate:endDate]
self._lastBaseHistList = histList
histList['p_change'] = histList['close'].pct_change().round(5) * 100
if indexs:
return self.getData(indexs=indexs)
else:
return histList
elif autype == 'qfq':
bfqData = histList.assign(date=pd.to_datetime(histList['datetime'].apply(lambda x: str(x[0:10])))).assign(code=str(code))\
.assign(date_stamp=histList['datetime'].apply(lambda x: self._dateStamp(str(x)[0:10])))
bfqData = bfqData.set_index('date')
bfqData = bfqData.drop(
['year', 'month', 'day', 'hour', 'minute', 'datetime'], axis=1)
xdxrData = self.getXdxr(code)
if xdxrData is not None:
info = xdxrData[xdxrData['category'] == 1]
bfqData['if_trade'] = True
data = pd.concat([bfqData, info[['category']]
[bfqData.index[0]:]], axis=1)
#data['date'] = data.index
data['if_trade'].fillna(value=False, inplace=True)
data = data.fillna(method='ffill')
data = pd.concat([data, info[['fenhong', 'peigu', 'peigujia',
'songzhuangu']][bfqData.index[0]:]], axis=1)
data = data.fillna(0)
data['preclose'] = (data['close'].shift(1) * 10 - data['fenhong'] + data['peigu']
* data['peigujia']) / (10 + data['peigu'] + data['songzhuangu'])
data['adj'] = (data['preclose'].shift(-1) /
data['close']).fillna(1)[::-1].cumprod()
data['open'] = data['open'] * data['adj']
data['high'] = data['high'] * data['adj']
data['low'] = data['low'] * data['adj']
data['close'] = data['close'] * data['adj']
data['preclose'] = data['preclose'] * data['adj']
data = data[data['if_trade']]
histList = data.drop(['fenhong', 'peigu', 'peigujia', 'songzhuangu', 'if_trade', 'category', 'preclose', 'date_stamp', 'adj'], axis=1)
histList = histList[startDate:endDate]
self._lastBaseHistList = histList
histList['p_change'] = histList['close'].pct_change().round(5) * 100
if indexs:
return self.getData(indexs=indexs)
else:
return histList
else:
bfqData['preclose'] = bfqData['close'].shift(1)
bfqData['adj'] = 1
histList = bfqData.drop(['preclose', 'date_stamp', 'adj'], axis=1)
histList = histList[startDate:endDate]
self._lastBaseHistList = histList
if indexs:
return self.getData(indexs=indexs)
else:
return histList
elif autype == 'hfq':
xdxrData = self.getXdxr(code)
info = xdxrData[xdxrData['category'] == 1]
bfqData = histList.assign(date=histList['datetime'].apply(lambda x: x[0:10])).assign(code=str(code))\
.assign(date_stamp=histList['datetime'].apply(lambda x: self._dateStamp(str(x)[0:10])))
bfqData = bfqData.set_index('date')
bfqData = bfqData.drop(
['year', 'month', 'day', 'hour', 'minute', 'datetime'], axis=1)
bfqData['if_trade'] = True
data = pd.concat([bfqData, info[['category']]
[bfqData.index[0]:]], axis=1)
data['if_trade'].fillna(value=False, inplace=True)
data = data.fillna(method='ffill')
data = pd.concat([data, info[['fenhong', 'peigu', 'peigujia',
'songzhuangu']][bfqData.index[0]:]], axis=1)
data = data.fillna(0)
data['preclose'] = (data['close'].shift(1) * 10 - data['fenhong'] + data['peigu']
* data['peigujia']) / (10 + data['peigu'] + data['songzhuangu'])
data['adj'] = (data['preclose'].shift(-1) /
data['close']).fillna(1).cumprod()
data['open'] = data['open'] / data['adj']
data['high'] = data['high'] / data['adj']
data['low'] = data['low'] / data['adj']
data['close'] = data['close'] / data['adj']
data['preclose'] = data['preclose'] / data['adj']
data = data[data['if_trade']]
histList = data.drop(['fenhong', 'peigu', 'peigujia', 'songzhuangu', 'if_trade', 'category', 'preclose', 'date_stamp', 'adj'], axis=1)
histList = histList[startDate:endDate]
self._lastBaseHistList = histList
histList['p_change'] = histList['close'].pct_change().round(5) * 100
if indexs:
return self.getData(indexs=indexs)
else:
return histList
#按分钟得到标准数据
'''
ktype = 1/5/15/30/60 分钟
'''
def getMins(self, code, ktype = 1, start = '', end = '', indexs=['vol', 'ma', 'macd', 'kdj', 'cci', 'bbi', 'sar', 'trix']):
startDate, endDate, diffDayNum = self._getDate(start, end)
self._connect()
ktypeCode = 8
if int(ktype) == 1:
ktypeCode = 8
diffDayNum = 240 * diffDayNum
elif int(ktype) == 5:
ktypeCode = 0
diffDayNum = 48 * diffDayNum
elif int(ktype) == 15:
ktypeCode = 1
diffDayNum = 16 * diffDayNum
elif int(ktype) == 30:
ktypeCode = 2
diffDayNum = 8 * diffDayNum
elif int(ktype) == 60:
ktypeCode = 3
diffDayNum = 4 * diffDayNum
if diffDayNum > 20800:
diffDayNum = 20800
histList = pd.concat([self.__tdx.to_df(self.__tdx.get_security_bars(ktypeCode, self.getMarketCode(
str(code)), str(code), (int(diffDayNum / 800) - i) * 800, 800)) for i in range(int(diffDayNum / 800) + 1)], axis=0)
if histList.empty:
return None
histList = histList\
.assign(datetime=pd.to_datetime(histList['datetime']), code=str(code))\
.assign(date=histList['datetime'].apply(lambda x: str(x)[0:10]))\
.assign(date_stamp=histList['datetime'].apply(lambda x: self._dateStamp(x)))\
.assign(time_stamp=histList['datetime'].apply(lambda x: self._timeStamp(x)))
histList['date'] = histList['datetime']
histList = histList.drop(['year', 'month', 'day', 'hour', 'minute', 'datetime', 'date_stamp', 'time_stamp'], axis=1)
histList = histList.set_index('date')
histList = histList[startDate:endDate]
self._lastBaseHistList = histList
histList['p_change'] = histList['close'].pct_change().round(5) * 100
histList['vol'] = histList['vol'] / 100.0
if indexs:
return self.getData(indexs=indexs)
else:
return histList
#按天得到指数日k线
'''
ktype = D(天)/W(周)/M(月)/Q(季)/Y(年)
'''
def getIndexDays(self, code, ktype = 'D', start = '', end = '', indexs=['turnover', 'vol', 'ma', 'macd', 'kdj', 'cci', 'bbi', 'sar', 'trix']):
startDate, endDate, diffDayNum = self._getDate(start, end)
self._connect()
ktypeCode = 9
if ktype.lower() == 'd':
ktypeCode = 9
elif ktype.lower() == 'w':
ktypeCode = 5
elif ktype.lower() == 'm':
ktypeCode = 6
elif ktype.lower() == 'q':
ktypeCode = 10
elif ktype.lower() == 'y':
ktypeCode = 11
if str(code)[0] in ['5', '1']: # ETF
data = pd.concat([self.__tdx.to_df(self.__tdx.get_security_bars(
ktypeCode, 1 if str(code)[0] in ['0', '8', '9', '5'] else 0, code, (int(diffDayNum / 800) - i) * 800, 800)) for i in range(int(diffDayNum / 800) + 1)], axis=0)
else:
data = pd.concat([self.__tdx.to_df(self.__tdx.get_index_bars(
ktypeCode, 1 if str(code)[0] in ['0', '8', '9', '5'] else 0, code, (int(diffDayNum / 800) - i) * 800, 800)) for i in range(int(diffDayNum / 800) + 1)], axis=0)
histList = data.assign(date=data['datetime'].apply(lambda x: str(x[0:10]))).assign(code=str(code))\
.assign(date_stamp=data['datetime'].apply(lambda x: self._dateStamp(str(x)[0:10])))\
.assign(code=code)
if histList.empty:
return None
histList = histList.drop(['year', 'month', 'day', 'hour', 'minute', 'datetime', 'date_stamp', 'up_count', 'down_count'], axis=1)
histList = histList.set_index('date')
histList = histList[startDate:endDate]
self._lastBaseHistList = histList
histList['p_change'] = histList['close'].pct_change().round(5) * 100
if indexs:
return self.getData(indexs=indexs)
else:
return histList
#按分钟得到标准数据
'''
ktype = 1/5/15/30/60 分钟
'''
def getIndexMins(self, code, ktype = 1, start = '', end = '', indexs=['vol', 'ma', 'macd', 'kdj', 'cci', 'bbi', 'sar', 'trix']):
startDate, endDate, diffDayNum = self._getDate(start, end)
self._connect()
ktypeCode = 8
if int(ktype) == 1:
ktypeCode = 8
diffDayNum = 240 * diffDayNum
elif int(ktype) == 5:
ktypeCode = 0
diffDayNum = 48 * diffDayNum
elif int(ktype) == 15:
ktypeCode = 1
diffDayNum = 16 * diffDayNum
elif int(ktype) == 30:
ktypeCode = 2
diffDayNum = 8 * diffDayNum
elif int(ktype) == 60:
ktypeCode = 3
diffDayNum = 4 * diffDayNum
if diffDayNum > 20800:
diffDayNum = 20800
if str(code)[0] in ['5', '1']: # ETF
data = pd.concat([self.__tdx.to_df(self.__tdx.get_security_bars(
ktypeCode, 1 if str(code)[0] in ['0', '8', '9', '5'] else 0, code, (int(diffDayNum / 800) - i) * 800, 800)) for i in range(int(diffDayNum / 800) + 1)], axis=0)
else:
data = pd.concat([self.__tdx.to_df(self.__tdx.get_index_bars(
ktypeCode, 1 if str(code)[0] in ['0', '8', '9', '5'] else 0, code, (int(diffDayNum / 800) - i) * 800, 800)) for i in range(int(diffDayNum / 800) + 1)], axis=0)
histList = data.assign(datetime=pd.to_datetime(data['datetime']), code=str(code))\
.assign(date=data['datetime'].apply(lambda x: str(x)[0:10]))\
.assign(date_stamp=data['datetime'].apply(lambda x: self._dateStamp(x)))\
.assign(time_stamp=data['datetime'].apply(lambda x: self._timeStamp(x)))
if histList.empty:
return None
histList['date'] = histList['datetime']
histList = histList.drop(['year', 'month', 'day', 'hour', 'minute', 'datetime', 'date_stamp', 'time_stamp', 'up_count', 'down_count'], axis=1)
histList = histList.set_index('date')
histList = histList[startDate:endDate]
self._lastBaseHistList = histList
histList['p_change'] = histList['close'].pct_change().round(5) * 100
if indexs:
return self.getData(indexs=indexs)
else:
return histList
#实时逐笔
'''
0买 1卖 2中性
'''
def getRealtimeTransaction(self, code):
self._connect()
try:
data = pd.concat([self.__tdx.to_df(self.__tdx.get_transaction_data(
self.getMarketCode(str(code)), code, (2 - i) * 2000, 2000)) for i in range(3)], axis=0)
if 'value' in data.columns:
data = data.drop(['value'], axis=1)
data = data.dropna()
day = datetime.date.today()
histList = data.assign(date=str(day)).assign(datetime=pd.to_datetime(data['time'].apply(lambda x: str(day) + ' ' + str(x))))\
.assign(code=str(code)).assign(order=range(len(data.index)))
histList['money'] = histList['price'] * histList['vol'] * 100
histList['type'] = histList['buyorsell']
histList['type'].replace([0,1,2], ['B','S','N'], inplace = True)
histList = histList.drop(['order', 'buyorsell'], axis=1).reset_index()
return histList
except:
return None
#历史逐笔
'''
0买 1卖 2中性
'''
def getHistoryTransaction(self, code, date):
self._connect()
try:
data = pd.concat([self.__tdx.to_df(self.__tdx.get_history_transaction_data(
self.getMarketCode(str(code)), code, (2 - i) * 2000, 2000, int(str(date).replace('-', '')))) for i in range(3)], axis=0)
if 'value' in data.columns:
data = data.drop(['value'], axis=1)
data = data.dropna()
#day = datetime.date.today()
day = date
histList = data.assign(date=str(day)).assign(datetime=pd.to_datetime(data['time'].apply(lambda x: str(day) + ' ' + str(x))))\
.assign(code=str(code)).assign(order=range(len(data.index)))
histList['money'] = histList['price'] * histList['vol'] * 100
histList['type'] = histList['buyorsell']
histList['type'].replace([0,1,2], ['B','S','N'], inplace = True)
histList = histList.drop(['order', 'buyorsell'], axis=1).reset_index()
return histList
except:
return None
#实时分时数据
def getRealtimeMinuteTime(self, code):
self._connect()
date = str(time.strftime('%Y-%m-%d',time.localtime()))
morningData = pd.date_range(start=str(date) + ' 09:31', end=str(date) + ' 11:30', freq = 'min')
morningDF = pd.DataFrame(index=morningData)
afternoonData = pd.date_range(start=str(date) + ' 13:01',end=str(date) + ' 15:00', freq = 'min')
afternoonDF = pd.DataFrame(index=afternoonData)
timeData = morningDF.append(afternoonDF)
histList = self.__tdx.to_df(self.__tdx.get_minute_time_data(
self.getMarketCode(str(code)), code))
#非标准均价计算
money = histList['price'] * histList['vol'] * 100
histList['money'] = money.round(2)
totalMoney = money.cumsum()
totalVol = histList['vol'].cumsum()
histList['ave'] = totalMoney / (totalVol * 100)
histList['ave'] = histList['ave'].round(3)
histList['datetime'] = timeData.index[0:len(histList)]
histList['date'] = histList['datetime'].apply(lambda x: x.strftime('%Y-%m-%d'))
histList['time'] = histList['datetime'].apply(lambda x: x.strftime('%H:%M'))
histList = histList.reset_index()
return histList
#历史分时数据
def getHistoryMinuteTime(self, code, date):
self._connect()
morningData = pd.date_range(start=str(date) + ' 09:31', end=str(date) + ' 11:30', freq = 'min')
morningDF = pd.DataFrame(index=morningData)
afternoonData = pd.date_range(start=str(date) + ' 13:01',end=str(date) + ' 15:00', freq = 'min')
afternoonDF = pd.DataFrame(index=afternoonData)
timeData = morningDF.append(afternoonDF)
histList = self.__tdx.to_df(self.__tdx.get_history_minute_time_data(
self.getMarketCode(str(code)), code, int(str(date).replace('-', ''))))
#非标准均价计算
money = histList['price'] * histList['vol'] * 100
histList['money'] = money.round(2)
totalMoney = money.cumsum()
totalVol = histList['vol'].cumsum()
histList['ave'] = totalMoney / (totalVol * 100)
histList['ave'] = histList['ave'].round(3)
histList['datetime'] = timeData.index[0:len(histList)]
histList['date'] = histList['datetime'].apply(lambda x: x.strftime('%Y-%m-%d'))
histList['time'] = histList['datetime'].apply(lambda x: x.strftime('%H:%M'))
histList = histList.reset_index()
return histList
#实时报价(五档行情)
'''
market => 市场
active1 => 活跃度
price => 现价
last_close => 昨收
open => 开盘
high => 最高
low => 最低
reversed_bytes0 => 保留
reversed_bytes1 => 保留
vol => 总量
cur_vol => 现量
amount => 总金额
s_vol => 内盘
b_vol => 外盘
reversed_bytes2 => 保留
reversed_bytes3 => 保留
bid1 => 买一价
ask1 => 卖一价
bid_vol1 => 买一量
ask_vol1 => 卖一量
bid2 => 买二价
ask2 => 卖二价
bid_vol2 => 买二量
ask_vol2 => 卖二量
bid3 => 买三价
ask3 => 卖三价
bid_vol3 => 买三量
ask_vol3 => 卖三量
bid4 => 买四价
ask4 => 卖四价
bid_vol4 => 买四量
ask_vol4 => 卖四量
bid5 => 买五价
ask5 => 卖五价
bid_vol5 => 买五量
ask_vol5 => 卖五量
reversed_bytes4 => 保留
reversed_bytes5 => 保留
reversed_bytes6 => 保留
reversed_bytes7 => 保留
reversed_bytes8 => 保留
reversed_bytes9 => 涨速
active2 => 活跃度
'''
def getRealtimeQuotes(self, codeList):
self._connect()
itemList = []
for item in codeList:
itemList.append((self.getMarketCode(item), item))
histList = self.__tdx.to_df(self.__tdx.get_security_quotes(itemList))
histList = histList.set_index('code')
return histList
#计算指定日期成交量细节
def getVolAnalysis(self, code, date):
self._connect()
if str(time.strftime('%Y-%m-%d',time.localtime())) == str(date):
if int(time.strftime('%H%M',time.localtime())) > 1600:
volList = self.getHistoryTransaction(code, date)
else:
volList = self.getRealtimeTransaction(code)
else:
volList = self.getHistoryTransaction(code, date)
if volList is None:
return None
guben,circulate = self.getGuben(code)
if not self._lastBaseHistList.empty:
histList = self._lastBaseHistList.copy()
else:
histList = self.getDays(code, end=date, indexs=[])
#涨停单数量
limitVol = round(histList[-5:]['vol'].mean() * 0.0618)
#超大单,先转成市值,再转回成手数
superVol = float(circulate) * float(histList['close'].values[-1]) * 0.000618 / float(histList['close'].values[-1]) / 100
#大单
bigVol = round(superVol * 0.518)
#中单
middleVol = round(superVol * 0.382)
#小单
smallVol = round(superVol * 0.191)
#买单统计
buyVolList = volList[volList['type'] == 'B']
totalBuyVolNum = buyVolList['vol'].sum()
mainBuyVolNum = buyVolList[buyVolList['vol'] >= bigVol]['vol'].sum()
limitBuyVolNum = math.ceil(buyVolList[(buyVolList['vol'] >= limitVol)]['vol'].sum() / limitVol)
superBuyVolNum = math.ceil(buyVolList[(buyVolList['vol'] < limitVol) & (buyVolList['vol'] >= superVol)]['vol'].sum() / superVol)
bigBuyVolNum = math.ceil(buyVolList[(buyVolList['vol'] < superVol) & (buyVolList['vol'] >= bigVol)]['vol'].sum() / bigVol)
middleBuyVolNum = math.ceil(buyVolList[(buyVolList['vol'] < bigVol) & (buyVolList['vol'] >= middleVol)]['vol'].sum() / middleVol)
smallBuyVolNum = math.ceil(buyVolList[(buyVolList['vol'] < middleVol) & (buyVolList['vol'] >= smallVol)]['vol'].sum() / smallVol)
microBuyVolNum = len(buyVolList[(buyVolList['vol'] < smallVol)])
#print limitBuyVolNum,superBuyVolNum,bigBuyVolNum,middleBuyVolNum,smallBuyVolNum,microBuyVolNum
#卖单统计
sellVolList = volList[volList['type'] == 'S']
totalSellVolNum = sellVolList['vol'].sum()
mainSellVolNum = sellVolList[sellVolList['vol'] >= bigVol]['vol'].sum()
limitSellVolNum = math.ceil(sellVolList[(sellVolList['vol'] >= limitVol)]['vol'].sum() / limitVol)
superSellVolNum = math.ceil(sellVolList[(sellVolList['vol'] < limitVol) & (sellVolList['vol'] >= superVol)]['vol'].sum() / superVol)
bigSellVolNum = math.ceil(sellVolList[(sellVolList['vol'] < superVol) & (sellVolList['vol'] >= bigVol)]['vol'].sum() / bigVol)
middleSellVolNum = math.ceil(sellVolList[(sellVolList['vol'] < bigVol) & (sellVolList['vol'] >= middleVol)]['vol'].sum() / middleVol)
smallSellVolNum = math.ceil(sellVolList[(sellVolList['vol'] < middleVol) & (sellVolList['vol'] >= smallVol)]['vol'].sum() / smallVol)
microSellVolNum = len(sellVolList[(sellVolList['vol'] < smallVol)])
#print limitSellVolNum,superSellVolNum,bigSellVolNum,middleSellVolNum,smallSellVolNum,microSellVolNum
#计算吸筹线
#主力标准吸筹金额
mainBaseMoney = round(histList['close'].values[-1] * circulate * 0.001 / 10000 / 10000, 4)
#主力强力吸筹金额
mainBigMoney = round(histList['close'].values[-1] * circulate * 0.003 / 10000 / 10000, 4)
#资金统计
totalMoney = round(volList['money'].sum() / 10000 / 10000, 4)
totalBuyMoney = round(buyVolList['money'].sum() / 10000 / 10000, 4)
totalSellMoney = round(sellVolList['money'].sum() / 10000 / 10000, 4)
totalAbsMoney = round(totalBuyMoney - totalSellMoney, 3)
mainMoney = round(volList[volList['vol'] >= bigVol]['money'].sum() / 10000 / 10000, 4)
mainBuyMoney = round(buyVolList[buyVolList['vol'] >= bigVol]['money'].sum() / 10000 / 10000, 4)
mainSellMoney = round(sellVolList[sellVolList['vol'] >= bigVol]['money'].sum() / 10000 / 10000, 4)
mainAbsMoney = round(mainBuyMoney - mainSellMoney, 3)
mainRate = 0
try:
mainRate = round((mainBuyMoney + mainSellMoney) / totalMoney * 100, 2)
except:
pass
mainBuyRate = 0
try:
mainBuyRate = round(mainBuyMoney / (mainBuyMoney + mainSellMoney) * 100, 2)
except:
pass
#print totalAbsMoney,mainAbsMoney,totalMoney,totalBuyMoney,totalSellMoney,mainBuyMoney,mainSellMoney,mainRate,mainBuyRate
#成交笔数
volNum = len(volList)
#平均每笔交易价格
aveTradePrice = round(totalMoney / volNum * 10000 * 10000, 2)
#平均每股买价格
avePerShareBuyPrice = 0
try:
avePerShareBuyPrice = round(totalBuyMoney * 10000 * 10000 / (totalBuyVolNum * 100), 3)
except:
pass
#主力平均每股买价格
mainAvePerShareBuyPrice = 0
try:
mainAvePerShareBuyPrice = round(mainBuyMoney * 10000 * 10000 / (mainBuyVolNum * 100), 3)
except:
pass
#平均每股卖价格
avePerShareSellPrice = 0
try:
avePerShareSellPrice = round(totalSellMoney * 10000 * 10000 / (totalSellVolNum * 100), 3)
except:
pass
#主力平均每股卖价格
mainAvePerShareSellPrice = 0
try:
mainAvePerShareSellPrice = round(mainSellMoney * 10000 * 10000 / (mainSellVolNum * 100), 3)
except:
pass
#print totalMoney,volNum,aveVolPrice * 10000 * 10000
statData = {}
statData['limit_buy_vol_num'] = limitBuyVolNum
statData['super_buy_vol_num'] = superBuyVolNum
statData['big_buy_vol_num'] = bigBuyVolNum
statData['middle_buy_vol_num'] = middleBuyVolNum
statData['small_buy_vol_num'] = smallBuyVolNum
statData['micro_buy_vol_num'] = microBuyVolNum
statData['limit_sell_vol_num'] = limitSellVolNum
statData['super_sell_vol_num'] = superSellVolNum
statData['big_sell_vol_num'] = bigSellVolNum
statData['middle_sell_vol_num'] = middleSellVolNum
statData['small_sell_vol_num'] = smallSellVolNum
statData['micro_sell_vol_num'] = microSellVolNum
statData['total_abs_money'] = totalAbsMoney
statData['main_abs_money'] = mainAbsMoney
statData['total_money'] = totalMoney
statData['total_buy_money'] = totalBuyMoney
statData['total_sell_money'] = totalSellMoney
statData['main_money'] = mainMoney
statData['main_buy_money'] = mainBuyMoney
statData['main_sell_money'] = mainSellMoney
statData['main_rate'] = mainRate
statData['main_buy_rate'] = mainBuyRate
statData['trade_num'] = volNum
statData['vol_num'] = volList['vol'].sum()
statData['ave_trade_price'] = aveTradePrice
statData['main_base_money'] = mainBaseMoney
statData['main_big_money'] = mainBigMoney
statData['ave_per_share_buy_price'] = avePerShareBuyPrice
statData['ave_per_share_sell_price'] = avePerShareSellPrice
statData['main_ave_per_share_buy_price'] = mainAvePerShareBuyPrice
statData['main_ave_per_share_sell_price'] = mainAvePerShareSellPrice
statData['circulate_money'] = round(circulate * histList['close'].values[-1] / 10000 / 10000, 4)
return statData
#输出ebk文件
def outputEbk(self, stockList, ebkPath = ''):
if len(ebkPath) <= 0:
ebkPath = os.getcwd() + '/' + sys.argv[0][0:-3] + '.' + str(time.strftime('%Y%m%d',time.localtime())) + '.ebk'
if not isinstance(stockList,list):
return False
fp = open(ebkPath, "a")
fp.write('\r\n') #ebk第一行为空行
for code in stockList:
if self.getMarketCode(code) == 1:
fp.write('1' + code)
else:
fp.write('0' + code)
fp.write('\r\n')
fp.close()
return True
#输出sel文件
def outputSel(self, stockList, selPath = ''):
import struct
if len(selPath) <= 0:
selPath = os.getcwd() + '/' + sys.argv[0][0:-3] + '.' + str(time.strftime('%Y%m%d',time.localtime())) + '.sel'
if not isinstance(stockList,list):
return False
stocks = []
for code in stockList:
if self.getMarketCode(code) == 1:
stocks.append('\x07\x11' + code)
else:
stocks.append('\x07\x21' + code)
with open(selPath, 'ab') as fp:
data = struct.pack('H', len(stocks)).decode() + ''.join(stocks)
fp.write(data.encode())
return True
#ebk to sel
def ebk2sel(self, ebkPath):
import struct
if not os.path.exists(ebkPath):
return False
selPath = ebkPath.replace('.ebk', '.sel')
stocks = []
with open(ebkPath, 'r') as ebkfp:
for code in ebkfp:
code = code.strip()
if len(code) > 0:
if self.getMarketCode(code[1:]) == 1:
stocks.append('\x07\x11' + code[1:])
else:
stocks.append('\x07\x21' + code[1:])
with open(selPath, 'wb') as selfp:
data = struct.pack('H', len(stocks)).decode() + ''.join(stocks)
selfp.write(data.encode())
return True
#sel to ebk
def sel2ebk(self, selPath):
import struct
if not os.path.exists(selPath):
return False
ebkPath = selPath.replace('.sel', '.ebk')
with open(selPath, 'rb') as selfp:
ebkfp = open(ebkPath, "a")
cnt = struct.unpack('<H', selfp.read(2))[0]
for _ in range(cnt):
data = selfp.readline(8).decode()
exch = '1' if data[1] == '\x11' else '0'
code = exch + data[2:]
ebkfp.write(code + '\r\n')
ebkfp.close()
return True
class MongoDB(object):
def __init__(
self,
ip="stock_mongo", #mongo db 数据库docker 容器名
port=27017,
user_name=None,
pwd=None,
authdb=None):
self.__server_ip = ip
self.__server_port = port
self.__user_name = user_name
self.__pwd = pwd
self.__authdb = authdb
self.client = None
self.trade_day = True
self.TDX_IP_SETS = STOCK_IP_SETS
self.api = TdxHq_API(heartbeat=True)
self.today = None
self.accounts = []
self.db = "stock_mock" #数据库
self.account_collection = "account" #保存各个账户的当前资金信息
self.account_his_collection = "account_history" #保存每个账户的历史净值信息
self.prefix = "holdlist_"
self.accounts = []
self.stocks = []
def connect(self):
'''建立数据库的连接
'''
_db_session = MongoClient(self.__server_ip, self.__server_port)
if self.__user_name:
eval("_db_session.{}".format(self.__authdb)).authenticate(
self.__user_name, self.__pwd)
self.client = _db_session
def connect_market(self):
for ip in self.TDX_IP_SETS:
try:
if self.api.connect(ip, 7709):
return
except:
pass
def disconnect(self):
'''断开数据库连接
'''
self.client.close()
return True
def _dbclient(self, db):
'''返回某个特定数据库的对象
'''
return eval("self.client.{}".format(db))
def handle_ex_right(self):
'''处理持仓股票除权价格和数量
'''
func = lambda x: 0 if not x else x
today = datetime.datetime.today().date().day
year = datetime.datetime.today().date().year
month = datetime.datetime.today().date().month
for stock in self.stocks:
mk = self._select_market_code(stock)
cqcx = self.api.get_xdxr_info(mk, stock)[::-1]
dct = {"fenhong": 0, 'peigu': 0, 'peigujia': 0, "songzhuangu": 0}
iscq = False
for i in cqcx:
if i["day"] != today or i["month"] != month or i[
"year"] != year:
break
else:
iscq = True
dct["fenhong"] += func(i["fenhong"])
dct["peigu"] += func(i["peigu"])
dct["peigujia"] += func(i["peigujia"])
dct["songzhuangu"] += func(i["songzhuangu"])
if iscq: #发生除权除息
rst = self.api.get_security_bars(4, mk, stock, 0, 2)
if rst[0]["day"] != today or i["month"] != month or i[
"year"] != year:
close = rst[0]["close"]
else:
close = rst[1]["close"]
preclose = (close * 10 - dct["fenhong"] +
dct["peigu"] * dct['peigujia']) / (
10 + dct['peigu'] + dct['songzhuangu'])
rate = close / preclose
logger.info("除权除息：{}，rate：{}".format(stock, rate))
for account in self.accounts:
filt = {"code": stock, "cx_date": {"$ne": self.today}}
dt = {
"$mul": {
"cost": 1 / rate,
"number": rate
},
"$set": {
"cx_date": self.today
}
}
self._dbclient(self.db)[self.prefix + account].update_one(
filt, dt)
def set_accounts(self):
self.accounts = [
i["account"]
for i in self._dbclient(self.db)[self.account_collection].find()
]
def set_stocks(self):
rst = []
for account in self.accounts:
rst.extend([
i["code"]
for i in self._dbclient(self.db)[self.prefix + account].find(
{"number": {
"$gt": 0
}}, {
"_id": 0,
"code": 1
})
])
self.stocks = set(rst)
def initial(self):
'''每天初始化状态，连接行情数据源，更新除权信息
'''
self.today = datetime.datetime.today().date().strftime('%Y-%m-%d')
df = ts.trade_cal()
self.trade_day = df[(
df["calendarDate"] == self.today)].isOpen.values[0]
if self.trade_day: #交易日，连接数据库，连接行情源，处理除权除息
self.connect()
self.connect_market()
self.set_accounts()
self.set_stocks()
self.handle_ex_right()
logger.info("initial finished")
def _select_market_code(self, code):
code = str(code)
if code[0] in ['5', '6', '9'] or code[:3] in [
"009", "126", "110", "201", "202", "203", "204"
]:
return 1
return 0
def updateaccount(self, account="test"):
'''更新账户净值，添加一条净值记录
'''
hold_collection = self.prefix + account
tm = datetime.datetime.now().strftime("%Y-%m-%d %H:%M:%S")
plimit = 80
filt = {'number': {'$gt': 0}}
stocks = [(self._select_market_code(i["code"]), i["code"])
for i in self._dbclient(self.db)[hold_collection].find(filt)]
rst = {}
for i in range(math.ceil(len(stocks) / plimit)):
ss = stocks[i * plimit:(i + 1) * plimit]
for item in self.api.get_security_quotes(ss):
rst[item["code"]] = item["price"]
if len(rst) > 0:
bulk = self._dbclient(
self.db)[hold_collection].initialize_ordered_bulk_op()
for code, price in rst.items():
d = {"price": price, "update_datetime": tm}
bulk.find({"code": code}).upsert().update({"$set": d})
bulk.execute()
holdvalue = self._dbclient(self.db)[hold_collection].aggregate([{
"$group": {
"_id": None,
"total": {
"$sum": {
"$multiply": ["$price", "$number"]
}
}
}
}])
try:
holdvalue = [i for i in holdvalue][0]["total"]
except:
holdvalue = 0
rst = self._dbclient(self.db)[self.account_collection].find(
{"account": account}, {"_id": 0})[0]
rst["total"] = rst["rest"] + holdvalue
rst["hold"] = holdvalue
#更新账户当前值
self._dbclient(self.db)[self.account_collection].update_one(
{"account": account}, {"$set": rst})
#更新账户历史记录值
dtm = datetime.datetime.now()
minute = dtm.minute
if minute >= 50:
dtm = dtm.replace(hour=dtm.hour + 1, minute=0)
else:
minute = int(minute / 10) * 10 + 10
dtm = dtm.replace(minute=minute)
self._dbclient(self.db)[self.account_his_collection].update_one(
{
"account": account,
"date": dtm.strftime("%Y-%m-%d %H:%M")
}, {"$set": rst},
upsert=True)
def run(self):
if not self.trade_day: #交易日，更新账户信息
return
for account in self.accounts:
self.updateaccount(account)
def QA_fetch_get_stock_day(code, start_date, end_date, if_fq='00', frequence='day', ip=best_ip['stock'], port=7709):
"""获取日线及以上级别的数据
Arguments:
code {str:6} -- code 是一个单独的code 6位长度的str
start_date {str:10} -- 10位长度的日期 比如'2017-01-01'
end_date {str:10} -- 10位长度的日期 比如'2018-01-01'
Keyword Arguments:
if_fq {str} -- '00'/'bfq' -- 不复权 '01'/'qfq' -- 前复权 '02'/'hfq' -- 后复权 '03'/'ddqfq' -- 定点前复权 '04'/'ddhfq' --定点后复权
frequency {str} -- day/week/month/quarter/year 也可以是简写 D/W/M/Q/Y
ip {str} -- [description] (default: best_ip['stock']) ip可以通过select_best_ip()函数重新获取
port {int} -- [description] (default: {7709})
Returns:
pd.DataFrame/None -- 返回的是dataframe,如果出错比如只获取了一天,而当天停牌,返回None
Exception:
如果出现网络问题/服务器拒绝, 会出现socket:time out 尝试再次获取/更换ip即可, 本函数不做处理
"""
api = TdxHq_API()
with api.connect(ip, port, time_out=0.7):
if frequence in ['day', 'd', 'D', 'DAY', 'Day']:
frequence = 9
elif frequence in ['w', 'W', 'Week', 'week']:
frequence = 5
elif frequence in ['month', 'M', 'm', 'Month']:
frequence = 6
elif frequence in ['quarter', 'Q', 'Quarter', 'q']:
frequence = 10
elif frequence in ['y', 'Y', 'year', 'Year']:
frequence = 11
start_date = str(start_date)[0:10]
today_ = datetime.date.today()
lens = QA_util_get_trade_gap(start_date, today_)
data = pd.concat([api.to_df(api.get_security_bars(frequence, _select_market_code(
code), code, (int(lens / 800) - i) * 800, 800)) for i in range(int(lens / 800) + 1)], axis=0)
# 这里的问题是: 如果只取了一天的股票,而当天停牌, 那么就直接返回None了
if len(data) < 1:
return None
data = data[data['open'] != 0]
if if_fq in ['00', 'bfq']:
data = data.assign(date=data['datetime'].apply(lambda x: str(x[0:10]))).assign(code=str(code))\
.assign(date_stamp=data['datetime'].apply(lambda x: QA_util_date_stamp(str(x)[0:10]))).set_index('date', drop=False, inplace=False)
return data.drop(['year', 'month', 'day', 'hour', 'minute', 'datetime'], axis=1)[start_date:end_date].assign(date=data['date'].apply(lambda x: str(x)[0:10]))
elif if_fq in ['01', 'qfq']:
xdxr_data = QA_fetch_get_stock_xdxr(code)
bfq_data = data.assign(date=pd.to_datetime(data['datetime'].apply(lambda x: str(x[0:10])))).assign(code=str(code))\
.assign(date_stamp=data['datetime'].apply(lambda x: QA_util_date_stamp(str(x)[0:10]))).set_index('date', drop=False, inplace=False)
bfq_data = bfq_data.drop(
['year', 'month', 'day', 'hour', 'minute', 'datetime'], axis=1)
#
if xdxr_data is not None:
info = xdxr_data[xdxr_data['category'] == 1]
bfq_data['if_trade'] = True
data = pd.concat([bfq_data, info[['category']]
[bfq_data.index[0]:]], axis=1)
data['date'] = data.index
data['if_trade'].fillna(value=False, inplace=True)
data = data.fillna(method='ffill')
data = pd.concat([data, info[['fenhong', 'peigu', 'peigujia',
'songzhuangu']][bfq_data.index[0]:]], axis=1)
data = data.fillna(0)
data['preclose'] = (data['close'].shift(1) * 10 - data['fenhong'] + data['peigu']
* data['peigujia']) / (10 + data['peigu'] + data['songzhuangu'])
data['adj'] = (data['preclose'].shift(-1) /
data['close']).fillna(1)[::-1].cumprod()
data['open'] = data['open'] * data['adj']
data['high'] = data['high'] * data['adj']
data['low'] = data['low'] * data['adj']
data['close'] = data['close'] * data['adj']
data['preclose'] = data['preclose'] * data['adj']
data = data[data['if_trade']]
return data.drop(['fenhong', 'peigu', 'peigujia', 'songzhuangu', 'if_trade', 'category'], axis=1)[data['open'] != 0].assign(date=data['date'].apply(lambda x: str(x)[0:10]))[start_date:end_date]
else:
bfq_data['preclose'] = bfq_data['close'].shift(1)
bfq_data['adj'] = 1
return bfq_data[start_date:end_date]
elif if_fq in ['03', 'ddqfq']:
xdxr_data = QA_fetch_get_stock_xdxr(code)
info = xdxr_data[xdxr_data['category'] == 1]
bfq_data = data\
.assign(date=pd.to_datetime(data['datetime'].apply(lambda x: x[0:10])))\
.assign(code=str(code))\
.assign(date_stamp=data['datetime'].apply(lambda x: QA_util_date_stamp(str(x)[0:10])))\
.set_index('date', drop=False, inplace=False)\
.drop(['year', 'month', 'day', 'hour',
'minute', 'datetime'], axis=1)
bfq_data['if_trade'] = True
data = pd.concat([bfq_data, info[['category']]
[bfq_data.index[0]:end_date]], axis=1)
data['date'] = data.index
data['if_trade'].fillna(value=False, inplace=True)
data = data.fillna(method='ffill')
data = pd.concat([data, info[['fenhong', 'peigu', 'peigujia',
'songzhuangu']][bfq_data.index[0]:end_date]], axis=1)
data = data.fillna(0)
data['preclose'] = (data['close'].shift(1) * 10 - data['fenhong'] + data['peigu']
* data['peigujia']) / (10 + data['peigu'] + data['songzhuangu'])
data['adj'] = (data['preclose'].shift(-1) /
data['close']).fillna(1)[::-1].cumprod()
data['open'] = data['open'] * data['adj']
data['high'] = data['high'] * data['adj']
data['low'] = data['low'] * data['adj']
data['close'] = data['close'] * data['adj']
data['preclose'] = data['preclose'] * data['adj']
data = data[data['if_trade']]
return data.drop(['fenhong', 'peigu', 'peigujia', 'songzhuangu', 'if_trade', 'category'], axis=1)[data['open'] != 0].assign(date=data['date'].apply(lambda x: str(x)[0:10]))[start_date:end_date]
elif if_fq in ['02', 'hfq']:
xdxr_data = QA_fetch_get_stock_xdxr(code)
info = xdxr_data[xdxr_data['category'] == 1]
bfq_data = data\
.assign(date=pd.to_datetime(data['datetime'].apply(lambda x: x[0:10])))\
.assign(code=str(code))\
.assign(date_stamp=data['datetime'].apply(lambda x: QA_util_date_stamp(str(x)[0:10])))\
.set_index('date', drop=False, inplace=False)\
.drop(['year', 'month', 'day', 'hour',
'minute', 'datetime'], axis=1)
bfq_data['if_trade'] = True
data = pd.concat([bfq_data, info[['category']]
[bfq_data.index[0]:]], axis=1)
data['date'] = data.index
data['if_trade'].fillna(value=False, inplace=True)
data = data.fillna(method='ffill')
data = pd.concat([data, info[['fenhong', 'peigu', 'peigujia',
'songzhuangu']][bfq_data.index[0]:]], axis=1)
data = data.fillna(0)
data['preclose'] = (data['close'].shift(1) * 10 - data['fenhong'] + data['peigu']
* data['peigujia']) / (10 + data['peigu'] + data['songzhuangu'])
data['adj'] = (data['preclose'].shift(-1) /
data['close']).fillna(1).cumprod()
data['open'] = data['open'] / data['adj']
data['high'] = data['high'] / data['adj']
data['low'] = data['low'] / data['adj']
data['close'] = data['close'] / data['adj']
data['preclose'] = data['preclose'] / data['adj']
data = data[data['if_trade']]
return data.drop(['fenhong', 'peigu', 'peigujia', 'songzhuangu', 'if_trade', 'category'], axis=1)[data['open'] != 0].assign(date=data['date'].apply(lambda x: str(x)[0:10]))[start_date:end_date]
elif if_fq in ['04', 'ddhfq']:
xdxr_data = QA_fetch_get_stock_xdxr(code)
info = xdxr_data[xdxr_data['category'] == 1]
bfq_data = data\
.assign(date=pd.to_datetime(data['datetime'].apply(lambda x: x[0:10])))\
.assign(code=str(code))\
.assign(date_stamp=data['datetime'].apply(lambda x: QA_util_date_stamp(str(x)[0:10])))\
.set_index('date', drop=False, inplace=False)\
.drop(['year', 'month', 'day', 'hour',
'minute', 'datetime'], axis=1)
bfq_data['if_trade'] = True
data = pd.concat([bfq_data, info[['category']]
[bfq_data.index[0]:end_date]], axis=1)
data['date'] = data.index
data['if_trade'].fillna(value=False, inplace=True)
data = data.fillna(method='ffill')
data = pd.concat([data, info[['fenhong', 'peigu', 'peigujia',
'songzhuangu']][bfq_data.index[0]:end_date]], axis=1)
data = data.fillna(0)
data['preclose'] = (data['close'].shift(1) * 10 - data['fenhong'] + data['peigu']
* data['peigujia']) / (10 + data['peigu'] + data['songzhuangu'])
data['adj'] = (data['preclose'].shift(-1) /
data['close']).fillna(1).cumprod()
data['open'] = data['open'] / data['adj']
data['high'] = data['high'] / data['adj']
data['low'] = data['low'] / data['adj']
data['close'] = data['close'] / data['adj']
data['preclose'] = data['preclose'] / data['adj']
data = data[data['if_trade']]
return data.drop(['fenhong', 'peigu', 'peigujia', 'songzhuangu', 'if_trade', 'category'], axis=1)[data['open'] != 0].assign(date=data['date'].apply(lambda x: str(x)[0:10]))[start_date:end_date]
