如何对不同选股方案进行组合？

步骤

组合方案以交集为例

Step_1： 读取选股结果的Excel

Step_2： 初始化admin，并计算交集的结果

Step_3： 计算交集的绩效

Step_4： 可视化收益绩效

读取选股结果的Excel

• 获取PB与PS选股的结果
• 获取MA选股的结果
• 将选股结果处理成MultiIndex格式

import pandas as pd
factor_opt = pd.read_excel('factor_opt.xlsx', index_col="date").stack()
selector_opt = pd.read_excel('selector_opt.xlsx', index_col="date").stack()

Step_2： 初始化admin，并计算交集的结果

• 输入两个优化选股结果表格
• 用Intersection_Strategy返回交集的结果

from fxdayu_alphaman.selector.admin import Admin
admin = Admin()
selector_name_list = ['factor_opt', 'selector_opt']
selector_result_list = [factor_opt, selector_opt]
Intersection = admin.Intersection_Strategy(dict(zip(selector_name_list,selector_result_list)))

Step_3： 计算交集的绩效

• 用calculate_performance计算策略交集方案的绩效

from datetime import datetime
intersection_perf = admin.calculate_performance('intersection', Intersection.strategy_result, start=datetime(2016,8,1), end=datetime(2017,4,20), 
                            periods=(5,10))
intersection_perf.key_performance_indicator

{'period_10': Annual return           0.392246
 Cumulative returns      0.410649
 Annual volatility       0.677822
 Sharpe ratio            0.806029
 Calmar ratio            0.964976
 Stability               0.124383
 Max drawdown           -0.406482
 Omega ratio             1.289168
 Sortino ratio           1.379652
 Skew                    3.561108
 Kurtosis               56.521566
 Tail ratio              1.109034
 Daily value at risk    -0.083229
 dtype: float64, 'period_5': Annual return           0.393688
 Cumulative returns      0.412168
 Annual volatility       0.617618
 Sharpe ratio            0.832133
 Calmar ratio            0.746088
 Stability               0.002849
 Max drawdown           -0.527669
 Omega ratio             1.355400
 Sortino ratio           1.484751
 Skew                    3.068953
 Kurtosis               52.906192
 Tail ratio              0.946620
 Daily value at risk    -0.075773
 dtype: float64}

from fxdayu_alphaman.selector.selector_analysis import *
plot_cumulative_returns(intersection_perf.mean_return, period=10)
plt.show()

Intersection.strategy_result.unstack().replace(np.nan,0).replace(2,1).to_excel('intersection.xlsx')