def _update_k_data(self, code, k_data):
# 注意：所有的数据库数据和列表数据都按照日期的正序排序(从小到大)
"""
更新股票，股指每日数据（行情，K线，市值等0）
@:param code 股票(指数)代码
@:param k_data ts中获取的最新df数据
"""
if len(k_data) != 0:
k_data = k_data.sort_values(by='trade_date')
cl_stock_code = self.db[code]
cl_stock_code.create_index([('trade_date', ASCENDING)], unique=True)
# 更新k线数据
# 1、新增日K线入库
# 2、遍历数据库找出最近的500+22(必须保证更新数据操作在22天以内进行)条数据并更新最后的22条的ma和最高 最低价
for ix, row in k_data.iterrows():
d = row.to_dict()
d['ts_code'] = d['ts_code'].replace('.', '_')
if self._is_in_vnpy_db(d['ts_code'], update=True):
# 更新vnpy数据库数据
self._build_db_vnpy(d)
flt = {'trade_date': d['trade_date']}
cl_stock_code.replace_one(flt, d, upsert=True)
rec = list(cl_stock_code.find({}).sort("trade_date", DESCENDING).limit(522))
rec.reverse()
am = ArrayManager(size=600)
last_5_vol = deque([0.0] * 5)
last_5_amount = deque([0.0] * 5)
for d in rec:
if 0.0 not in last_5_vol:
vol_rate = d['vol'] / (sum(last_5_vol) / 5.0)
amount_rate = d['amount'] / (sum(last_5_amount) / 5.0)
d['vol_rate'] = vol_rate
d['amount_rate'] = amount_rate
else:
d['vol_rate'] = 0.0
d['amount_rate'] = 0.0
last_5_vol.popleft()
last_5_amount.popleft()
last_5_vol.append(d['vol'])
last_5_amount.append(d['amount'])
if d['ts_code'][-3:] == '_SH':
exchange = Exchange.SSE
d['exchange'] = 'SSE'
else:
exchange = Exchange.SZSE
d['exchange'] = 'SZSE'
bar = BarData(
gateway_name='ctp', symbol=d['ts_code'],
exchange=exchange,
datetime=string_to_datetime(d['trade_date']))
bar.symbol = d['ts_code']
bar.volume = d['vol']
bar.open_price = d['open']
bar.high_price = d['high']
bar.low_price = d['low']
bar.close_price = d['close']
am.update_bar(bar)
if d['trade_date'] >= self.db_date:
d['ma_5'] = am.sma(5)
d['ma_10'] = am.sma(10)
d['ma_20'] = am.sma(20)
d['ma_30'] = am.sma(30)
d['ma_60'] = am.sma(60)
d['ma_120'] = am.sma(120)
d['ma_250'] = am.sma(250)
d['ma_500'] = am.sma(500)
d['high_5'] = np.max(am.high[-5:])
d['high_10'] = np.max(am.high[-10:])
d['high_20'] = np.max(am.high[-20:])
d['high_30'] = np.max(am.high[-30:])
d['high_60'] = np.max(am.high[-60:])
d['high_120'] = np.max(am.high[-120:])
d['high_250'] = np.max(am.high[-250:])
d['high_500'] = np.max(am.high[-500:])
d['low_5'] = np.min(am.low[-5:])
d['low_10'] = np.min(am.low[-10:])
d['low_20'] = np.min(am.low[-20:])
d['low_30'] = np.min(am.low[-30:])
d['low_60'] = np.min(am.low[-60:])
d['low_120'] = np.min(am.low[-120:])
d['low_250'] = np.min(am.low[-250:])
d['low_500'] = np.min(am.low[-500:])
flt = {'trade_date': d['trade_date']}
cl_stock_code.replace_one(flt, d, upsert=True)
def _init_k_data(self, code, k_data):
# 注意：所有的数据库数据和列表数据都按照日期的正序排序(从小到大)
"""
初始化股票数据库数据
@:param code 股票(指数)代码
"""
if len(k_data) != 0:
last_5_vol = deque([0.0] * 5)
last_5_amount = deque([0.0] * 5)
k_data = k_data.sort_values(by='trade_date')
cl_stock_code = self.db[code]
cl_stock_code.create_index([('trade_date', ASCENDING)], unique=True)
am = ArrayManager(size=600)
for ix, row in k_data.iterrows():
d = row.to_dict()
d['ts_code'] = d['ts_code'].replace('.', '_')
if 0.0 not in last_5_vol:
vol_rate = d['vol'] / (sum(last_5_vol) / 5.0)
amount_rate = d['amount'] / (sum(last_5_amount) / 5.0)
d['vol_rate'] = vol_rate
d['amount_rate'] = amount_rate
else:
d['vol_rate'] = 0.0
d['amount_rate'] = 0.0
last_5_vol.popleft()
last_5_amount.popleft()
last_5_vol.append(d['vol'])
last_5_amount.append(d['amount'])
if self._is_in_vnpy_db(d['ts_code'], update=False):
# 构建vnpy股票数据库数据
self._build_db_vnpy(d)
if d['ts_code'][-3:] == '_SH':
exchange = Exchange.SSE
d['exchange'] = 'SSE'
else:
exchange = Exchange.SZSE
d['exchange'] = 'SZSE'
bar = BarData(
gateway_name='ctp', symbol=d['ts_code'],
exchange=exchange,
datetime=string_to_datetime(d['trade_date']))
bar.symbol = d['ts_code']
bar.volume = d['vol']
bar.open_price = d['open']
bar.high_price = d['high']
bar.low_price = d['low']
bar.close_price = d['close']
am.update_bar(bar)
try:
d['ma_5'] = am.sma(5)
except:
traceback.print_exc()
LOG.error('************************')
LOG.error(d['ts_code'])
LOG.error(d['trade_date'])
LOG.error(bar)
d['ma_10'] = am.sma(10)
d['ma_20'] = am.sma(20)
d['ma_30'] = am.sma(30)
d['ma_60'] = am.sma(60)
d['ma_120'] = am.sma(120)
d['ma_250'] = am.sma(250)
d['ma_500'] = am.sma(500)
d['high_5'] = np.max(am.high[-5:])
d['high_10'] = np.max(am.high[-10:])
d['high_20'] = np.max(am.high[-20:])
d['high_30'] = np.max(am.high[-30:])
d['high_60'] = np.max(am.high[-60:])
d['high_120'] = np.max(am.high[-120:])
d['high_250'] = np.max(am.high[-250:])
d['high_500'] = np.max(am.high[-500:])
d['low_5'] = np.min(am.low[-5:])
d['low_10'] = np.min(am.low[-10:])
d['low_20'] = np.min(am.low[-20:])
d['low_30'] = np.min(am.low[-30:])
d['low_60'] = np.min(am.low[-60:])
d['low_120'] = np.min(am.low[-120:])
d['low_250'] = np.min(am.low[-250:])
d['low_500'] = np.min(am.low[-500:])
flt = {'trade_date': d['trade_date']}
cl_stock_code.replace_one(flt, d, upsert=True)