def __init__(self):
super(FundDection, self).__init__('../log/FundDetection.log')
self.engine = self.get_engine()
self.db_session = self.get_session()
self.sess = self.db_session()
self.ts_util = TushareBaseUtil()
class FundDection(BaseService):
def __init__(self):
super(FundDection, self).__init__('../log/FundDetection.log')
self.engine = self.get_engine()
self.db_session = self.get_session()
self.sess = self.db_session()
self.ts_util = TushareBaseUtil()
def get_engine(self):
return DBSelector().get_engine('db_stock')
def create_table(self):
# 初始化数据库连接:
Base.metadata.create_all(self.engine) # 创建表结构
def get_session(self):
return sessionmaker(bind=self.engine)
def start(self):
query_result_str, has_data = self.query_big_volatility_share()
if has_data:
title = f'{self.today} LOF/ETF 申购波动数据'
print(title)
print(query_result_str)
send_from_aliyun(title, content=query_result_str)
else:
self.logger.info(f'今天{self.today}没有数据')
def query_big_volatility_share(self):
day = 0 # 前x天 ，天 ，0 即使昨天和前天的数据比较
PERCENT = 10 # 偏离 百分比
DIFF_MAX = 200
category_list = ["LOF", "ETF"]
yesterday = self.ts_util.get_last_trade_date() # 最新的一天
yesterday = self.ts_util.date_convertor(yesterday)
lastday_of_yesterday = self.ts_util.get_trade_date()[-3 - day] # 上一天
lastday_of_yesterday = self.ts_util.date_convertor(
lastday_of_yesterday)
string_arg = ''
has_data = False
for category in category_list:
string_arg += f'\n############ {category} ###############\n\n'
lastest_lofs = self.sess.query(
FundBaseInfoModel.name, FundBaseInfoModel.code,
ShareModel.share, ShareModel.date).join(ShareModel).filter(
FundBaseInfoModel.category == category).filter(
or_(ShareModel.date == yesterday,
ShareModel.date == lastday_of_yesterday)).all()
current_df = pd.DataFrame(lastest_lofs)
current_df['date'] = current_df['date'].astype(str)
current_df['share'] = current_df['share'].astype(float)
current_df['code'] = current_df['code'].astype(str)
current_df['name'] = current_df['name'].astype(str)
for code, sub_df in current_df.groupby('code'):
# 按照code分组，把前后两个值做差值
yesterday_share = sub_df[sub_df['date'] == yesterday]
lastday_of_yesterday_share = sub_df[sub_df['date'] ==
lastday_of_yesterday]
if len(yesterday_share) > 0 and len(
lastday_of_yesterday_share) > 0:
yesterday_share_num = yesterday_share['share'].to_list()[0]
lastday_of_yesterday_num = lastday_of_yesterday_share[
'share'].to_list()[0]
diff_part = yesterday_share_num - lastday_of_yesterday_num
diff = (
diff_part) * 1.00 / lastday_of_yesterday_num * 100.00
diff = round(diff, 2)
if abs(diff) >= PERCENT or abs(diff_part) > DIFF_MAX:
has_data = True # 有数据则发送邮件
print(yesterday_share['name'].to_list()[0],
yesterday_share['code'].to_list()[0],
yesterday_share_num, lastday_of_yesterday_num,
lastday_of_yesterday, diff, round(diff_part, 0))
string = self.formator(
category, yesterday_share['name'].to_list()[0],
yesterday_share['code'].to_list()[0],
yesterday_share_num, lastday_of_yesterday_num,
yesterday, diff, round(diff_part, 0))
# print(string)
string_arg += string + '\n'
string_arg += '\n'
# print(string_arg)
return string_arg, has_data
def formator(self, *args):
string = '{} {} {} {}万份 {}万份 {} {}% 多出了{}万份\n'.format(*args)
return string