def format(self, entity, df):
cols = list(df.columns)
str_cols = ['Title']
date_cols = [self.get_original_time_field()]
float_cols = list(set(cols) - set(str_cols) - set(date_cols))
for column in float_cols:
df[column] = df[column].apply(lambda x: to_float(x[0]))
df.rename(columns=cash_flow_map, inplace=True)
df.update(df.select_dtypes(include=[np.number]).fillna(0))
if 'timestamp' not in df.columns:
df['timestamp'] = pd.to_datetime(
df[self.get_original_time_field()])
elif not isinstance(df['timestamp'].dtypes, datetime):
df['timestamp'] = pd.to_datetime(df['timestamp'])
df['report_period'] = df['timestamp'].apply(
lambda x: to_report_period_type(x))
df['report_date'] = pd.to_datetime(df['timestamp'])
df['entity_id'] = entity.id
df['provider'] = self.provider.value
df['code'] = entity.code
df['name'] = entity.name
df['id'] = self.generate_domain_id(entity, df)
return df
def format(self, entity, df):
df['volume'] = df['BianDongShuLiang'].apply(lambda x: to_float(x))
df['change_pct'] = df['BianDongBiLi'].apply(lambda x: to_float(x))
df['holding_pct'] = df['BianDongHouChiGuBiLi'].apply(lambda x: to_float(x))
df.update(df.select_dtypes(include=[np.number]).fillna(0))
df['holder_name'] = df['GuDongMingCheng'].astype(str)
df['holder_name'] = df['holder_name'].apply(lambda x: x.replace('\n', '').replace('\r', ''))
if 'timestamp' not in df.columns:
df['timestamp'] = pd.to_datetime(df[self.get_original_time_field()])
elif not isinstance(df['timestamp'].dtypes, datetime):
df['timestamp'] = pd.to_datetime(df['timestamp'])
df['entity_id'] = entity.id
df['provider'] = self.provider.value
df['code'] = entity.code
df['id'] = self.generate_domain_id(entity, df)
return df
def format(self, entity, df):
df['rights_issues'] = df['ShiJiPeiGu'].apply(lambda x: to_float(x))
df['rights_issue_price'] = df['PeiGuJiaGe'].apply(
lambda x: to_float(x))
df['rights_raising_fund'] = df['ShiJiMuJi'].apply(
lambda x: to_float(x))
df.update(df.select_dtypes(include=[np.number]).fillna(0))
if 'timestamp' not in df.columns:
df['timestamp'] = pd.to_datetime(
df[self.get_original_time_field()])
elif not isinstance(df['timestamp'].dtypes, datetime):
df['timestamp'] = pd.to_datetime(df['timestamp'])
df['entity_id'] = entity.id
df['provider'] = self.provider.value
df['code'] = entity.code
df['id'] = self.generate_domain_id(entity, df)
return df
def format(self, entity, df):
df['volume'] = df['BianDongShuLiang'].apply(lambda x: to_float(x))
df['price'] = df['JiaoYiJunJia'].apply(lambda x: to_float(x))
df['holding'] = df['BianDongHouShuLiang'].apply(lambda x: to_float(x))
df.update(df.select_dtypes(include=[np.number]).fillna(0))
df['trading_person'] = df['BianDongRen'].astype(str)
df['trading_person'] = df['trading_person'].apply(
lambda x: x.replace('\n', '').replace('\r', ''))
df['trading_way'] = df['JiaoYiTuJing'].astype(str)
df['trading_way'] = df['trading_way'].apply(
lambda x: x.replace('\n', '').replace('\r', ''))
df['manager'] = df['GaoGuanMingCheng'].astype(str)
df['manager'] = df['manager'].apply(
lambda x: x.replace('\n', '').replace('\r', ''))
df['manager_position'] = df['GaoGuanZhiWei'].astype(str)
df['manager_position'] = df['manager_position'].apply(
lambda x: x.replace('\n', '').replace('\r', ''))
df['relationship_with_manager'] = df['GaoGuanGuanXi'].astype(str)
df['relationship_with_manager'] = df[
'relationship_with_manager'].apply(
lambda x: x.replace('\n', '').replace('\r', ''))
if 'timestamp' not in df.columns:
df['timestamp'] = pd.to_datetime(
df[self.get_original_time_field()])
elif not isinstance(df['timestamp'].dtypes, datetime):
df['timestamp'] = pd.to_datetime(df['timestamp'])
df['entity_id'] = entity.id
df['provider'] = self.provider.value
df['code'] = entity.code
df['id'] = self.generate_domain_id(entity, df)
return df
def format(self, entity, df):
# 分红总额
df['dividend_money'] = df['FenHongZongE'].apply(lambda x: to_float(x[1]))
# 新股
df['ipo_issues'] = df['XinGu'].apply(lambda x: to_float(x[1]))
# 增发
df['spo_issues'] = df['ZengFa'].apply(lambda x: to_float(x[1]))
# 配股
df['rights_issues'] = df['PeiFa'].apply(lambda x: to_float(x[1]))
df.update(df.select_dtypes(include=[np.number]).fillna(0))
if 'timestamp' not in df.columns:
df['timestamp'] = pd.to_datetime(df[self.get_original_time_field()])
elif not isinstance(df['timestamp'].dtypes, datetime):
df['timestamp'] = pd.to_datetime(df['timestamp'])
df['entity_id'] = entity.id
df['provider'] = self.provider.value
df['code'] = entity.code
df['id'] = self.generate_domain_id(entity, df)
return df
def numba_boost_up(klines):
kdatas = []
# TODO: ignore the last unfinished kdata now,could control it better if need
for result in klines[:-1]:
# "2000-01-28,1005.26,1012.56,1173.12,982.13,3023326,3075552000.00"
# time,open,close,high,low,volume,turnover
fields = result.split(',')
kdatas.append(
dict(timestamp=fields[0],
open=to_float(fields[1]),
close=to_float(fields[2]),
high=to_float(fields[3]),
low=to_float(fields[4]),
volume=to_float(fields[5]),
turnover=to_float(fields[6])))
return kdatas
async def record(self, entity, http_session, db_session, para):
start_point = time.time()
(ref_record, start, end, size, timestamps) = para
json_results = []
for timestamp in timestamps:
timestamp_str = to_time_str(timestamp)
url = self.url.format(timestamp_str)
async with http_session.get(
url, headers=DEFAULT_SH_SUMMARY_HEADER) as response:
if response.status != 200:
return
text = await response.text()
if text is None:
continue
results = demjson.decode(text[text.index("(") +
1:text.index(")")])['result']
result = [
result for result in results
if result['productType'] == '1'
]
if result and len(result) == 1:
result_json = result[0]
# 有些较老的数据不存在,默认设为0.0
json_results.append({
'timestamp':
timestamp,
'pe':
to_float(result_json['profitRate'], 0.0),
'total_value':
to_float(result_json['marketValue1'] + '亿', 0.0),
'total_tradable_vaule':
to_float(result_json['negotiableValue1'] + '亿', 0.0),
'volume':
to_float(result_json['trdVol1'] + '万', 0.0),
'turnover':
to_float(result_json['trdAmt1'] + '亿', 0.0),
'turnover_rate':
to_float(result_json['exchangeRate'], 0.0),
})
if len(json_results) > self.batch_size:
df = pd.DataFrame.from_records(json_results)
df['entity_id'] = entity.id
df['provider'] = Provider.Exchange.value
df['timestamp'] = pd.to_datetime(df['timestamp'])
df['name'] = '上证指数'
df = self.format(df)
return False, time.time() - start_point, (ref_record,
df)
if len(json_results) > 0:
df = pd.DataFrame.from_records(json_results)
df = self.format(df)
return False, time.time() - start_point, (ref_record, df)
return True, time.time() - start_point, None
