From fccb2110b25d783a8260cb358f5df0ede8a3fa76 Mon Sep 17 00:00:00 2001
From: zhaojie01 <zhaojie01@wealthgrow.cn>
Date: Thu, 24 Dec 2020 12:03:06 +0800
Subject: [PATCH] =?UTF-8?q?=E5=AE=97=E7=86=B9=E6=96=B0=E7=AE=97=E6=B3=95?=
=?UTF-8?q?=E6=8E=A8=E8=8D=90?=
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
---
app/service/data_service_v2_1.py | 3 +
app/service/portfolio_diagnose.py | 164 ++++++++++++++++++++----------
2 files changed, 111 insertions(+), 56 deletions(-)
diff --git a/app/service/data_service_v2_1.py b/app/service/data_service_v2_1.py
index 31a1495..0cf3364 100644
--- a/app/service/data_service_v2_1.py
+++ b/app/service/data_service_v2_1.py
@@ -105,6 +105,7 @@ class UserCustomerDataAdaptor:
product_df = product_df.drop_duplicates("fund_id")
user_customer_order_df = order_df.set_index('fund_id').join(product_df.set_index('fund_id')).reset_index()
+ user_customer_order_df["confirm_share_date"] = user_customer_order_df["confirm_share_date"].apply(lambda x: pd.to_datetime(x.date()))
self.start_date = user_customer_order_df["confirm_share_date"].min()
self.customer_real_name = user_customer_order_df["customer_name"].values[0]
self.ifa_real_name = user_customer_order_df["username"].values[0]
@@ -162,6 +163,8 @@ class UserCustomerDataAdaptor:
data = cur.fetchall()
cur_fund_performance_df = pd.DataFrame(list(data),
columns=['price_date', 'ret_1w', 'ret_cum_1m', 'ret_cum_6m', 'ret_cum_1y', 'ret_cum_ytd', 'ret_cum_incep'])
+ cur_fund_performance_df = cur_fund_performance_df[
+ cur_fund_performance_df["price_date"] <= self.end_date.date()]
self.all_fund_performance[cur_fund_id] = cur_fund_performance_df
cur_fund_nav_df["price_date"] = pd.to_datetime(cur_fund_nav_df["price_date"])
diff --git a/app/service/portfolio_diagnose.py b/app/service/portfolio_diagnose.py
index ad4984b..53d28fe 100644
--- a/app/service/portfolio_diagnose.py
+++ b/app/service/portfolio_diagnose.py
@@ -124,7 +124,7 @@ def choose_good_evaluation(evaluation):
if v1[0] > 1:
del evaluation[1]
- if v2[0] > 1 and float(v2[1].strip('%')) <= 60:
+ if (v2[0] > 1 and float(v2[1].strip('%')) <= 60) or math.isnan(float(v2[1].strip('%'))):
del evaluation[2]
if v3[0] > 1:
del evaluation[3]
@@ -145,7 +145,7 @@ def choose_bad_evaluation(evaluation):
if v1[0] < 2:
del evaluation[1]
- if v2[0] < 2:
+ if v2[0] < 2 or math.isnan(float(v2[1].strip('%'))):
del evaluation[2]
if v3[0] < 2:
del evaluation[3]
@@ -235,10 +235,10 @@ def get_tamp_fund():
Returns:
"""
- with TAMP_SQL(tamp_fund_engine) as tamp_fund:
- tamp_fund_session = tamp_fund.session
- sql = "SELECT id FROM tamp_fund_info WHERE id LIKE 'HF%'"
- cur = tamp_fund_session.execute(sql)
+ with TAMP_SQL(tamp_product_engine) as tamp_prod:
+ tamp_prod_session = tamp_prod.session
+ sql = "SELECT id FROM fund_info WHERE `status` = 1 and strategy!=7"
+ cur = tamp_prod_session.execute(sql)
data = cur.fetchall()
# df = pd.read_sql(sql, con)
df = pd.DataFrame(list(data), columns=['fund_id'])
@@ -408,10 +408,13 @@ class PortfolioDiagnose(object):
prod = get_tamp_nav(self.portfolio[0], self.start_date, invest_type=self.invest_type)
fund_info = get_fund_info(self.end_date, invest_type=self.invest_type)
- # while prod is None or prod.index[-1] - prod.index[0] < 0.6 * (self.end_date - self.start_date):
- while prod is None:
+ while prod is None or prod.index[-1] - prod.index[0] < 0.6 * (self.end_date - self.start_date):
+ # while prod is None:
# 获å–的净值表为空时首先考虑基金净值数æ®ä¸è¶³åŠå¹´ï¼ŒæŸ¥æ‰¾åŒä¸€åŸºé‡‘ç»ç†ä¸‹çš„相åŒäºŒçº§ç–略的基金ID作替æ¢
result = fund_info[fund_info['fund_id'] == self.portfolio[0]]
+ if result.empty:
+ break
+
manager = str(result['manager'].values)
strategy = result['substrategy'].values
@@ -512,7 +515,7 @@ class PortfolioDiagnose(object):
# 组åˆå†…已包å«çš„ç–ç•¥
# included_strategy = set()
# 按æ¯ç§åŸºé‡‘最少投资100w确定组åˆåŒ…å«çš„最大基金数é‡
- max_len = len(self.portfolio) - len(prod.columns)
+ max_len = min(len(self.portfolio) - len(prod.columns), self.invest_amount/1e6)
# 排å表内包å«çš„所有ç–ç•¥
# all_strategy = set(fund_rank['substrategy'].to_list())
@@ -527,45 +530,84 @@ class PortfolioDiagnose(object):
# add_strategy = all_strategy - included_strategy
add_risk = all_risk - included_risk
- # é历产å“æ± ï¼ŒæŽ¨è得分>80且与组åˆå†…其他基金相关度低于0.8çš„å±žäºŽå¾…æ·»åŠ ç–略的基金
- for proposal in tamp_fund['fund_id']:
+ candidate_funds = tamp_fund['fund_id'].to_list()
+ candidate_info = []
+ for proposal in candidate_funds:
if proposal in fund_rank['fund_id'].to_list() and proposal not in prod.columns:
proposal_z_score = search_rank(fund_rank, proposal, metric='z_score')
proposal_strategy = fund_rank[fund_rank['fund_id'] == proposal]['substrategy'].values[0]
- else:
- continue
+ proposal_risk = get_risk_level(proposal_strategy)
+ if proposal_z_score >= 60:
+ candidate_info.append((proposal, proposal_z_score, proposal_risk))
- if proposal_z_score > 60 and (get_risk_level(proposal_strategy) in add_risk or not add_risk):
- # if proposal_z_score > 80:
- proposal_nav = get_tamp_nav(proposal, self.start_date, invest_type=self.invest_type)
- # 忽略净值周期大于周更的产å“
- # if get_frequency(proposal_nav) <= 52:
- # continue
-
- self.freq_list.append(get_frequency(proposal_nav))
- proposal_nav = rename_col(proposal_nav, proposal)
-
- # 按最大周期进行é‡é‡‡æ ·ï¼Œè®¡ç®—新建组åˆçš„相关性
- prod_with_new_fund = pd.merge(prod, proposal_nav, how='outer', on='end_date').astype(float)
- prod_with_new_fund.sort_index(inplace=True)
- prod_with_new_fund.ffill(inplace=True)
-
- prod_with_new_fund = resample(prod_with_new_fund, get_trade_cal(), min(self.freq_list))
- self.new_correlation = cal_correlation(prod_with_new_fund)
- judge_correlation = self.new_correlation.fillna(0)
-
- if np.all(judge_correlation < 0.8):
- prod = prod_with_new_fund
- self.proposal_fund.append(proposal)
- max_len -= 1
- # add_strategy -= {proposal_strategy}
- add_risk -= {get_risk_level(proposal_strategy)}
- # if len(add_strategy) == 0 or max_len == 0:
- if max_len == 0:
- break
- else:
- # prod.drop(columns=proposal, inplace=True)
- self.freq_list.pop()
+ candidate_info.sort(key=lambda elem: elem[1], reverse=True)
+ # candidate_high_risk = [i[0] for i in list(filter(lambda x: x[2] == 'H', candidate_info))]
+ # candidate_median_risk = [i[0] for i in list(filter(lambda x: x[2] == 'M', candidate_info))]
+ # candidate_low_risk = [i[0] for i in list(filter(lambda x: x[2] == 'L', candidate_info))]
+ candidate_funds = [i[0] for i in candidate_info]
+
+ # é历产å“æ± ï¼ŒæŽ¨è得分>80且与组åˆå†…其他基金相关度低于0.8çš„å±žäºŽå¾…æ·»åŠ ç–略的基金
+ # for proposal in candidate_funds:
+ # proposal_strategy = fund_rank[fund_rank['fund_id'] == proposal]['substrategy'].values[0]
+ # if get_risk_level(proposal_strategy) in add_risk or not add_risk:
+ # # if proposal_z_score > 80:
+ # proposal_nav = get_tamp_nav(proposal, self.start_date, invest_type=self.invest_type)
+ # # 忽略净值周期大于周更的产å“
+ # # if get_frequency(proposal_nav) <= 52:
+ # # continue
+ #
+ # self.freq_list.append(get_frequency(proposal_nav))
+ # proposal_nav = rename_col(proposal_nav, proposal)
+ #
+ # # 按最大周期进行é‡é‡‡æ ·ï¼Œè®¡ç®—新建组åˆçš„相关性
+ # prod = pd.merge(prod, proposal_nav, how='outer', on='end_date').astype(float)
+ # prod.sort_index(inplace=True)
+ # prod.ffill(inplace=True)
+ # prod = resample(prod, get_trade_cal(), min(self.freq_list))
+ #
+ # self.new_correlation = cal_correlation(prod)
+ # judge_correlation = self.new_correlation.fillna(0)
+ #
+ # if np.all(judge_correlation < 0.8):
+ # self.proposal_fund.append(proposal)
+ # max_len -= 1
+ # # add_strategy -= {proposal_strategy}
+ # add_risk -= {get_risk_level(proposal_strategy)}
+ # # if len(add_strategy) == 0 or max_len == 0:
+ # if max_len == 0:
+ # break
+ # else:
+ # prod.drop(columns=proposal, inplace=True)
+ # é历产å“æ± ï¼ŒæŽ¨è得分>80且与组åˆå†…其他基金相关度低于0.8çš„å±žäºŽå¾…æ·»åŠ ç–略的基金
+
+ for proposal in candidate_funds:
+ proposal_nav = get_tamp_nav(proposal, self.start_date, invest_type=self.invest_type)
+ # 忽略净值周期大于周更的产å“
+ # if get_frequency(proposal_nav) <= 52:
+ # continue
+ self.freq_list.append(get_frequency(proposal_nav))
+ proposal_nav = rename_col(proposal_nav, proposal)
+
+ # 按最大周期进行é‡é‡‡æ ·ï¼Œè®¡ç®—新建组åˆçš„相关性
+ temp = pd.merge(prod, proposal_nav, how='outer', on='end_date').astype(float)
+ temp.sort_index(inplace=True)
+ temp.ffill(inplace=True)
+ temp = resample(temp, get_trade_cal(), min(self.freq_list))
+
+ self.new_correlation = cal_correlation(temp)
+ judge_correlation = self.new_correlation.fillna(0)
+
+ if np.all(judge_correlation < 0.8):
+ self.proposal_fund.append(proposal)
+ max_len -= 1
+ # add_strategy -= {proposal_strategy}
+ add_risk -= {get_risk_level(proposal_strategy)}
+ # if len(add_strategy) == 0 or max_len == 0:
+
+ prod = temp
+
+ if max_len == 0:
+ break
prod.dropna(how='all', inplace=True)
prod.fillna(method='bfill', inplace=True)
@@ -588,9 +630,9 @@ class PortfolioDiagnose(object):
print("é历产å“æ± èŽ·å–候选推è时间：", end3 - end2)
# propose_portfolio.to_csv('test_portfolio.csv', encoding='gbk')
- mu = expected_returns.mean_historical_return(self.propose_portfolio, frequency=min(self.freq_list))
+ mu = [search_rank(fund_rank, x, 'annual_return') for x in self.propose_portfolio.columns]
S = risk_models.sample_cov(self.propose_portfolio, frequency=min(self.freq_list))
- dd = expected_returns.drawdown_from_prices(self.propose_portfolio)
+ dd = [search_rank(fund_rank, x, 'max_drawdown') for x in self.propose_portfolio.columns]
# if self.client_type == 1:
# proposal_risk = [[x, get_risk_level(search_rank(fund_rank, x, metric='substrategy'))] for x in
@@ -607,20 +649,30 @@ class PortfolioDiagnose(object):
propose_risk_mapper[fund] = str(get_risk_level(search_rank(fund_rank, fund, metric='substrategy')))
if self.client_type == 1:
- risk_upper = {"L": 0.6, "M": 0.4, "H": 0.0}
- risk_lower = {"L": 0.6, "M": 0.4, "H": 0.0}
+ risk_upper = {"M": 0.4, "H": 0.0}
+ risk_lower = {"L": 0.6}
+ self.expect_return = 0.12
+ self.expect_drawdown = 0.03
elif self.client_type == 2:
- risk_upper = {"L": 0.5, "M": 0.3, "H": 0.2}
- risk_lower = {"L": 0.5, "M": 0.3, "H": 0.2}
+ risk_upper = {"H": 0.2}
+ risk_lower = {"L": 0.5, "M": 0.3}
+ self.expect_return = 0.15
+ self.expect_drawdown = 0.05
elif self.client_type == 3:
- risk_upper = {"L": 0.3, "M": 0.5, "H": 0.2}
- risk_lower = {"L": 0.3, "M": 0.5, "H": 0.2}
+ risk_upper = {"L": 0.3, "H": 0.3}
+ risk_lower = {"M": 0.4}
+ self.expect_return = 0.18
+ self.expect_drawdown = 0.08
elif self.client_type == 4:
- risk_upper = {"L": 0.3, "M": 0.4, "H": 0.3}
- risk_lower = {"L": 0.3, "M": 0.4, "H": 0.3}
+ risk_upper = {"L": 0.2, "M": 0.4}
+ risk_lower = {"H": 0.4}
+ self.expect_return = 0.15
+ self.expect_drawdown = 0.20
elif self.client_type == 5:
- risk_upper = {"L": 0.0, "M": 0.5, "H": 0.5}
- risk_lower = {"L": 0.0, "M": 0.5, "H": 0.5}
+ risk_upper = {"L": 0.0, "M": 0.4}
+ risk_lower = {"H": 0.6}
+ self.expect_return = 0.25
+ self.expect_drawdown = 0.15
else:
risk_upper = {"H": 1.0}
risk_lower = {"L": 0.0}
--
2.18.1