Appearance
使用范例
该范例所使用的组合数据来自于通过 rqalpha-plus 回测框架回测后所保存的回测结果。
python
import os
import pandas as pd
import numpy as np
import rqdatac
import pickle
from rqalpha.mod.rqalpha_mod_sys_analyser.plot import *
from rqdatac import *
import rqpattr
from rqpattr.api import performance_attribute
import matplotlib.pyplot as plt
import warnings
warnings.filterwarnings("ignore")
rqdatac.init()
#读取回测后保存的回测结果
path = os.getcwd()+'\\bts\\bts'
files = '000906.XSHG_factor1_liquidity_v2_asfc.pkl'
result = pd.read_pickle(os.path.join(path, files))
print('result')
#out:
# {'sys_analyser': {'summary': {'strategy_name': 'strategy',
# 'start_date': '2020-06-10',
# 'end_date': '2022-06-30',
# 'strategy_file': 'strategy.py',
# 'run_type': 'BACKTEST',
# 'starting_cash': 'STOCK:1000000000.0',
# 'STOCK': 1000000000.0,
# 'benchmark': '000906.XSHG',
# 'benchmark_symbol': '中证800',
# 'alpha': 0.04161847252000668,
# 'beta': 0.9689441949400958,
# 'sharpe': 0.42678333868827145,
# 'excess_sharpe': 0.3582160902725003,
# 'information_ratio': 0.3453132024893898,
# 'downside_risk': 0.1655438441069776,
# 'tracking_error': 0.12271422135189096,
# 'sortino': 0.5852185528297059,
# 'volatility': 0.22699885420999039,
# 'excess_volatility': 0.007730269334000854,
# 'excess_annual_volatility': 0.12271422135189096,
# 'max_drawdown': 0.3449066882507446,
# 'excess_max_drawdown': 0.16459229054214705,
# 'excess_returns': 0.07486020583373909,
# 'excess_annual_returns': 0.03712968826159724,
# 'var': 0.023450724628239612,
# 'total_value': 1203328714.4768782,
# 'cash': 8210.926878128295,
# 'total_returns': 0.20332871447687806,
# 'annualized_returns': 0.09798110622545697,
# 'unit_net_value': 1.203328714476878,
# 'units': 1000000000.0,
# 'benchmark_total_returns': 0.11694720785066992,
# 'benchmark_annualized_returns': 0.05744300314969153,
# 'excess_cum_returns': 0.08638150662620814,
# 'max_drawdown_duration': IndexRange(start=305, end=498, start_date=datetime.date(2021, 9, 7), end_date=datetime.date(2022, 6, 30)),
# 'max_drawdown_duration_start_date': '2021-09-07',
# 'max_drawdown_duration_end_date': '2022-06-30',
# 'max_drawdown_duration_days': 296,
# 'turnover': 19.746910020902284,
# 'avg_daily_turnover': 0.03878575014506964,
# 'excess_max_drawdown_duration': IndexRange(start=69, end=277, start_date=datetime.date(2020, 9, 17), end_date=datetime.date(2021, 7, 29)),
# 'excess_max_drawdown_duration_start_date': '2020-09-17',
# 'excess_max_drawdown_duration_end_date': '2021-07-29',
# 'excess_max_drawdown_duration_days': 315,
# 'weekly_alpha': 0.0407174667550259,
# 'weekly_beta': 0.9442199165766891,
# 'weekly_sharpe': 0.4328984550718106,
# 'weekly_sortino': 0.6233499038107769,
# 'weekly_information_ratio': 0.38383642721331795,
# 'weekly_tracking_error': 0.015232196824837712,
# 'weekly_max_drawdown': 0.2874243182734171},
# 'trades': datetime trading_datetime order_book_id \
# datetime
# 2020-06-10 15:00:00 2020-06-10 15:00:00 2020-06-10 15:00:00 000671.XSHE
# 2020-06-10 15:00:00 2020-06-10 15:00:00 2020-06-10 15:00:00 603806.XSHG
# 2020-06-10 15:00:00 2020-06-10 15:00:00 2020-06-10 15:00:00 002773.XSHE
# 2020-06-10 15:00:00 2020-06-10 15:00:00 2020-06-10 15:00:00 000988.XSHE
# 2020-06-10 15:00:00 2020-06-10 15:00:00 2020-06-10 15:00:00 601811.XSHG
# ... ... ... ...
# 2022-06-13 15:00:00 2022-06-13 15:00:00 2022-06-13 15:00:00 511220.XSHG
# 2022-06-16 15:00:00 2022-06-16 15:00:00 2022-06-16 15:00:00 511220.XSHG
# 2022-06-23 15:00:00 2022-06-23 15:00:00 2022-06-23 15:00:00 511220.XSHG
# 2022-06-29 15:00:00 2022-06-29 15:00:00 2022-06-29 15:00:00 511220.XSHG
# 2022-06-30 15:00:00 2022-06-30 15:00:00 2022-06-30 15:00:00 511220.XSHG
# symbol side position_effect exec_id tax \
# datetime
# 2020-06-10 15:00:00 阳光城 BUY OPEN 16710930284727 0.0
# 2020-06-10 15:00:00 福斯特 BUY OPEN 16710930284728 0.0
# 2020-06-10 15:00:00 康弘药业 BUY OPEN 16710930284729 0.0
# 2020-06-10 15:00:00 华工科技 BUY OPEN 16710930284730 0.0
# 2020-06-10 15:00:00 新华文轩 BUY OPEN 16710930284731 0.0
# ... ... ... ... ... ...
# 2022-06-13 15:00:00 海富通城投ETF BUY OPEN 16710930286411 0.0
# 2022-06-16 15:00:00 海富通城投ETF BUY OPEN 16710930286412 0.0
# 2022-06-23 15:00:00 海富通城投ETF BUY OPEN 16710930286413 0.0
# 2022-06-29 15:00:00 海富通城投ETF BUY OPEN 16710930286414 0.0
# 2022-06-30 15:00:00 海富通城投ETF BUY OPEN 16710930286415 0.0
# commission last_quantity last_price order_id \
# datetime
# 2020-06-10 15:00:00 54331.39440 26711600 6.780 16710930364951
# 2020-06-10 15:00:00 38711.06820 3062100 42.140 16710930364963
# 2020-06-10 15:00:00 28030.80870 2358900 39.610 16710930364972
# 2020-06-10 15:00:00 24600.08340 3723900 22.020 16710930364985
# 2020-06-10 15:00:00 22208.87760 7243600 10.220 16710930365008
# ... ... ... ... ...
# 2022-06-13 15:00:00 11.89320 400 99.110 16710930376692
# 2022-06-16 15:00:00 178.43040 6000 99.128 16710930376694
# 2022-06-23 15:00:00 5.00000 100 99.259 16710930376696
# 2022-06-29 15:00:00 1844.92842 62600 98.239 16710930376698
# 2022-06-30 15:00:00 8.84205 300 98.245 16710930376700
# transaction_cost
# datetime
# 2020-06-10 15:00:00 54331.39440
# 2020-06-10 15:00:00 38711.06820
# 2020-06-10 15:00:00 28030.80870
# 2020-06-10 15:00:00 24600.08340
# 2020-06-10 15:00:00 22208.87760
# ... ...
# 2022-06-13 15:00:00 11.89320
# 2022-06-16 15:00:00 178.43040
# 2022-06-23 15:00:00 5.00000
# 2022-06-29 15:00:00 1844.92842
# 2022-06-30 15:00:00 8.84205
# [1689 rows x 13 columns],
# 'portfolio': cash total_value market_value unit_net_value \
# date
# 2020-06-10 153.0189 9.997001e+08 9.996999e+08 0.999700
# 2020-06-11 153.0189 9.918064e+08 9.918062e+08 0.991806
# 2020-06-12 1984.0438 9.998441e+08 9.998421e+08 0.999844
# 2020-06-15 1984.0438 9.891545e+08 9.891525e+08 0.989154
# 2020-06-16 1984.0438 1.006052e+09 1.006050e+09 1.006052
# ... ... ... ... ...
# 2022-06-24 161.1973 1.187876e+09 1.187876e+09 1.187876
# 2022-06-27 161.1973 1.196714e+09 1.196679e+09 1.196714
# 2022-06-28 161.1973 1.205983e+09 1.205948e+09 1.205983
# 2022-06-29 2893.2689 1.168880e+09 1.168843e+09 1.168880
# 2022-06-30 8210.9269 1.203329e+09 1.203321e+09 1.203329
# units static_unit_net_value
# date
# 2020-06-10 1.000000e+09 1.0000
# 2020-06-11 1.000000e+09 0.9997
# 2020-06-12 1.000000e+09 0.9918
# 2020-06-15 1.000000e+09 0.9998
# 2020-06-16 1.000000e+09 0.9892
# ... ... ...
# 2022-06-24 1.000000e+09 1.1671
# 2022-06-27 1.000000e+09 1.1879
# 2022-06-28 1.000000e+09 1.1967
# 2022-06-29 1.000000e+09 1.2060
# 2022-06-30 1.000000e+09 1.1689
# [499 rows x 6 columns],
# 'benchmark_portfolio': unit_net_value
# date
# 2020-06-10 0.998862
# 2020-06-11 0.989584
# 2020-06-12 0.991023
# 2020-06-15 0.982023
# 2020-06-16 0.997713
# ... ...
# 2022-06-24 1.096298
# 2022-06-27 1.107309
# 2022-06-28 1.119947
# 2022-06-29 1.101584
# 2022-06-30 1.116947
# [499 rows x 1 columns],
# 'stock_account': cash transaction_cost market_value total_value
# date
# 2020-06-10 153.0189 299909.9811 9.996999e+08 9.997001e+08
# 2020-06-11 153.0189 0.0000 9.918062e+08 9.918064e+08
# 2020-06-12 1984.0438 1853.5725 9.998421e+08 9.998441e+08
# 2020-06-15 1984.0438 0.0000 9.891525e+08 9.891545e+08
# 2020-06-16 1984.0438 0.0000 1.006050e+09 1.006052e+09
# ... ... ... ... ...
# 2022-06-24 161.1973 0.0000 1.187876e+09 1.187876e+09
# 2022-06-27 161.1973 0.0000 1.196679e+09 1.196714e+09
# 2022-06-28 161.1973 0.0000 1.205948e+09 1.205983e+09
# 2022-06-29 2893.2689 1844.9284 1.168843e+09 1.168880e+09
# 2022-06-30 8210.9269 8.8421 1.203321e+09 1.203329e+09
# [499 rows x 4 columns],
# 'stock_positions': order_book_id symbol quantity last_price avg_price \
# date
# 2020-06-10 000671.XSHE 阳光城 26711600.0 6.78 6.780
# 2020-06-10 603806.XSHG 福斯特 3062100.0 42.14 42.140
# 2020-06-10 002773.XSHE 康弘药业 2358900.0 39.61 39.610
# 2020-06-10 000988.XSHE 华工科技 3723900.0 22.02 22.020
# 2020-06-10 601811.XSHG 新华文轩 7243600.0 10.22 10.220
# ... ... ... ... ... ...
# 2022-06-30 002460.XSHE 赣锋锂业 38100.0 148.70 141.450
# 2022-06-30 600316.XSHG 洪都航空 101100.0 30.24 31.603
# 2022-06-30 601117.XSHG 中国化学 300600.0 9.41 9.644
# 2022-06-30 600079.XSHG 人福医药 149200.0 16.00 18.070
# 2022-06-30 600588.XSHG 用友网络 42500.0 21.71 29.880
# market_value
# date
# 2020-06-10 181104648.0
# 2020-06-10 129036894.0
# 2020-06-10 93436029.0
# 2020-06-10 82000278.0
# 2020-06-10 74029592.0
# ... ...
# 2022-06-30 5665470.0
# 2022-06-30 3057264.0
# 2022-06-30 2828646.0
# 2022-06-30 2387200.0
# 2022-06-30 922675.0
# [19797 rows x 6 columns],
# 'yearly_risk_free_rates': {2020: 0.020009000000000002, 2021: 0, 2022: 0}}}获取回测结果中持仓及权益数据
python
positions = result['sys_analyser']['stock_positions']
portfolio = result['sys_analyser']['portfolio']
positions = positions.join(portfolio['total_value'])
positions['asset_type'] = 'stock'
positions = positions.reset_index()
positions = positions.set_index(['date','order_book_id','asset_type'])
daily_return = portfolio['unit_net_value'].pct_change().dropna()
weights = positions['market_value']/positions['total_value']
print('daily_return')
# out:
# date
# 2020-06-11 -0.007896
# 2020-06-12 0.008104
# 2020-06-15 -0.010692
# 2020-06-16 0.017083
# 2020-06-17 0.005524
# ...
# 2022-06-24 0.017760
# 2022-06-27 0.007440
# 2022-06-28 0.007745
# 2022-06-29 -0.030766
# 2022-06-30 0.029472
# Name: unit_net_value, Length: 498, dtype: float64
print('weights')
# out:
# date order_book_id asset_type
# 2020-06-10 000671.XSHE stock 0.181159
# 603806.XSHG stock 0.129076
# 002773.XSHE stock 0.093464
# 000988.XSHE stock 0.082025
# 601811.XSHG stock 0.074052
# ...
# 2022-06-30 002460.XSHE stock 0.004708
# 600316.XSHG stock 0.002541
# 601117.XSHG stock 0.002351
# 600079.XSHG stock 0.001984
# 600588.XSHG stock 0.000767
# Length: 19797, dtype: float64传入 rqpattr 计算引擎，对策略进行归因
python
result = rqpattr.performance_attribute(model="equity/factor_v2", daily_weights=weights, daily_return=daily_return,
benchmark_info={'type': 'index', 'name': '中证800', 'detail': '000906.XSHG'})混合资产 Brinson 归因
python
result['returns_decomposition']
#out:
# [{'factor': '总收益',
# 'value': 0.20369010703210916,
# 'children': [{'factor': '交易收益',
# 'value': -0.025843778212483993,
# 'children': None},
# {'factor': '杠杆收益', 'value': 0.0, 'children': None},
# {'factor': '持仓收益',
# 'value': 0.2295338852445934,
# 'children': [{'factor': '主动收益',
# 'value': 0.06632185697753212,
# 'children': [{'factor': '可转债收益',
# 'value': 0.0,
# 'children': [{'factor': '可转债配置收益', 'value': 0.0, 'children': None},
# {'factor': '可转债选择收益', 'value': 0.0, 'children': None}]},
# {'factor': '基金收益',
# 'value': 0.0,
# 'children': [{'factor': '基金配置收益', 'value': 0.0, 'children': None},
# {'factor': '基金选择收益', 'value': 0.0, 'children': None}]},
# {'factor': '指数收益',
# 'value': 0.0,
# 'children': [{'factor': '指数配置收益', 'value': 0.0, 'children': None},
# {'factor': '指数选择收益', 'value': 0.0, 'children': None}]},
# {'factor': '期权收益',
# 'value': 0.0,
# 'children': [{'factor': '期权配置收益', 'value': 0.0, 'children': None},
# {'factor': '期权选择收益', 'value': 0.0, 'children': None}]},
# {'factor': '期货收益',
# 'value': 0.0,
# 'children': [{'factor': '期货配置收益', 'value': 0.0, 'children': None},
# {'factor': '期货选择收益', 'value': 0.0, 'children': None}]},
# {'factor': '港股收益',
# 'value': 0.0,
# 'children': [{'factor': '港股配置收益', 'value': 0.0, 'children': None},
# {'factor': '港股选择收益', 'value': 0.0, 'children': None}]},
# {'factor': '现金收益',
# 'value': 0.0,
# 'children': [{'factor': '现金配置收益', 'value': 0.0, 'children': None},
# {'factor': '现金选择收益', 'value': 0.0, 'children': None}]},
# {'factor': '股票收益',
# 'value': 0.06632185697753212,
# 'children': [{'factor': '股票配置收益',
# 'value': 3.054056964001931e-06,
# 'children': None},
# {'factor': '股票选择收益',
# 'value': 0.06631880292056812,
# 'children': None}]},
# {'factor': '非标资产收益',
# 'value': 0.0,
# 'children': [{'factor': '非标资产配置收益', 'value': 0.0, 'children': None},
# {'factor': '非标资产选择收益', 'value': 0.0, 'children': None}]}]},
# {'factor': '基准持仓收益',
# 'value': 0.1632120282670611,
# 'children': [{'factor': '基准收益',
# 'value': 0.1182192041711989,
# 'children': None},
# {'factor': '穿透效应',
# 'value': 0.04499282409586221,
# 'children': None}]}]}]}]展示组合风格因子主动暴露及主动收益
python
style_factor_attr = pd.DataFrame(
result['attribution']['factor_attribution'][0]['factors'])
plt.figure()
style_factor_attr.set_index('factor')['active_exposure'].plot(kind='bar',figsize=(15,7),legend='active_exposure',fontsize=15)
plt.figure()
(style_factor_attr.set_index('factor')['active_return']*100).plot(kind='bar', figsize=(15, 7),legend='active_return',fontsize=15)
