def macroview_check(targetDate, markowitz_id='MZ.000070', weeks_lookback=52):
'''
ra_bl_view表 检查：宏观观点 最新数据是否更新
'''
print('\033[1;35m宏观观点################################## Check\033[0m')
gids = asset_mz_markowitz_asset.load([markowitz_id
])['mz_asset_id'].tolist()
# 获取ra_blview表格的宏观观点数据
db = database.connection('asset')
metadata = MetaData(bind=db)
t1 = Table('ra_bl_view', metadata, autoload=True)
columns = [
t1.c.globalid,
t1.c.bl_date,
t1.c.bl_index_id,
t1.c.bl_view,
]
s = select(columns)
startDate = (targetDate +
timedelta(weeks=-1 * weeks_lookback)).strftime('%Y-%m-%d')
s = s.where(t1.c.bl_date >= startDate).where(t1.c.globalid == 'BL.000001')
df = pd.read_sql(s,
db,
parse_dates=['bl_date'],
index_col=['bl_index_id', 'bl_date'])
gids.append('ALayer')
try:
for asset_id in gids:
v = df.loc[asset_id]
v = v.sort_index(ascending=True)
print(asset_id, v.index[-1], v.bl_view[-1])
print(
'markowitz配置资产宏观观点都存在，请检查观点是否正确：\033[1;35mPass inspection\033[0m')
return True
except:
print('markowitz配置资产宏观观点缺失：\033[1;43mFail inspection!!!\033[0m')
return False
def mz_asset_nav_check(targetDate, markowitz_id='MZ.000070'):
'''
markowitz --id MZ.000007 检查：MZ.000070配置指数及因子 最新数据是否更新
'''
print(
'\033[1;35mmarkowitz配置更新################################## Check\033[0m'
)
startDate = (targetDate + timedelta(weeks=-2)).strftime('%Y-%m-%d')
endDate = date.today().strftime('%Y-%m-%d')
assets_id = asset_mz_markowitz_asset.load([markowitz_id
])['mz_asset_id'].tolist()
assets = dict([(asset_id, Asset(asset_id).nav(begin_date=startDate,
end_date=endDate))
for asset_id in assets_id])
df_assets = pd.DataFrame(assets)
if df_assets.loc[targetDate].isnull().any():
print('markowitz配置资产或因子数据更新：\033[1;43mFail inspection!!!\033[0m')
print(df_assets.loc[[targetDate]])
return False
else:
print('markowitz配置资产或因子数据更新：\033[1;35mPass inspection\033[0m')
return True
def cov_check(targetDate, markowitz_id='MZ.000070'):
'''
on_online_cov表，检查：资产或因子 的最新数据是否完整且离谱
'''
print('\033[1;35m协方差#################################### Check\033[0m')
# 获取on_online_cov表格的数据
db = database.connection('asset')
metadata = MetaData(bind=db)
t1 = Table('on_online_cov', metadata, autoload=True)
columns = [
t1.c.on_asseta_id,
t1.c.on_assetb_id,
t1.c.on_date,
t1.c.on_cov,
]
s = select(columns)
s = s.where(t1.c.on_date == targetDate)
df = pd.read_sql(s, db, parse_dates=['on_date'])
gids = asset_mz_markowitz_asset.load([markowitz_id
])['mz_asset_id'].tolist()
if any(df) and df['on_cov'].isnull().any() == False:
print('协方差数据更新：\033[1;35mPass inspection\033[0m')
if len(df[df.on_asseta_id == df.on_assetb_id])**2 == len(df) and len(
df) == len(gids)**2:
print('协方差矩阵条数：\033[1;35mPass inspection\033[0m')
else:
print('协方差矩阵条数：\033[1;43mFail inspection!!!\033[0m')
df_covariance = df[df.on_asseta_id == df.on_assetb_id]
if (df_covariance['on_cov'] == 0.0).any():
print('方差非零：\033[1;43mFail inspection!!!\033[0m')
all_df = df_covariance[df_covariance['on_cov'] == 0].pivot(
index='on_date', columns='on_asseta_id', values='on_cov')
del all_df.index.name
print(all_df)
return False
else:
print('方差非零：\033[1;43mPass inspection!!!\033[0m')
df_covariance['stock'] = df_covariance['on_asseta_id'].map(
lambda x: int(x) if str(x)[:4] in ['1111'] else None)
df_covariance['bond'] = df_covariance['on_asseta_id'].map(
lambda x: int(x) if str(x)[:4] in ['1121'] else None)
stock_cov = df_covariance[~(
df_covariance['stock'].isnull())]['on_cov'].tolist()
bond_cov = df_covariance[~(
df_covariance['bond'].isnull())]['on_cov'].tolist()
if max(stock_cov) > min(stock_cov) * 4:
print('国内股票方差检查：\033[1;43mFail inspection!!!\033[0m')
stock_df = df_covariance[~(df_covariance['stock'].isnull())].pivot(
index='on_date', columns='on_asseta_id', values='on_cov')
del stock_df.index.name
print(stock_df)
else:
print('国内股票方差检查：\033[1;35mPass inspection\033[0m')
if max(bond_cov) > min(bond_cov) * 4:
print('国内债券方差正常：\033[1;43mFail inspection!!!\033[0m')
bond_df = df_covariance[~(df_covariance['bond'].isnull())].pivot(
index='on_date', columns='on_asseta_id', values='on_cov')
del bond_df.index.name
print(bond_df)
else:
print('国内债券方差正常：\033[1;35mPass inspection\033[0m')
else:
print('协方差数据更新：\033[1;43mFail inspection!!!\033[0m')
return True