1
开始选择策略
2
    │
3
    ├─ 第一步：排除不合格策略
4
    │   ├─ 总收益 < 5% → ❌ 淘汰
5
    │   ├─ 最大回撤 > 15% → ❌ 淘汰
6
    │   ├─ 胜率 < 50% → ❌ 淘汰
7
    │   └─ Sharpe < 1.0 → ❌ 淘汰
8
    │
9
    ├─ 第二步：根据目标分类
10
    │   ├─ 追求高收益 → 按总收益排序
11
    │   ├─ 追求低风险 → 按最大回撤排序
12
    │   └─ 追求平衡 → 按综合评分排序
13
    │
14
    ├─ 第三步：验证样本量
15
    │   ├─ 交易次数 < 10 → ⚠️ 样本不足，需更长测试
16
    │   ├─ 交易次数 10-50 → ✅ 合格
17
    │   └─ 交易次数 > 150 → ⚠️ 过度交易，检查手续费
18
    │
19
    └─ 第四步：最终确认
20
        ├─ 检查退出原因分布
21
        ├─ 检查持仓时间分布
22
        ├─ 进行样本外测试
23
        └─ 选定策略 ✅

1
#!/usr/bin/env python3
2
# -*- coding: utf-8 -*-
3

4
"""
5
策略自动选择工具
6
根据回测结果自动评分和排序
7
"""
8

9
strategies = [
10
    {
11
        "name": "MomentumTrendStrategy",
12
        "trades": 68,
13
        "avg_profit": 1.52,
14
        "total_profit": 23.45,
15
        "win_rate": 89.7,
16
        "max_drawdown": 3.21,
17
        "sharpe": 3.5
18
    },
19
    {
20
        "name": "Strategy003",
21
        "trades": 42,
22
        "avg_profit": 1.31,
23
        "total_profit": 18.72,
24
        "win_rate": 85.7,
25
        "max_drawdown": 4.15,
26
        "sharpe": 3.2
27
    },
28
    {
29
        "name": "Strategy001",
30
        "trades": 127,
31
        "avg_profit": 0.85,
32
        "total_profit": 15.33,
33
        "win_rate": 72.4,
34
        "max_drawdown": 6.82,
35
        "sharpe": 2.1
36
    }
37
]
38

39
def calculate_score(strategy):
40
    """计算综合评分"""
41
    # 收益分 (30%)
42
    profit_score = min(strategy["total_profit"] / 30 * 100, 100) * 0.30
43

44
    # 风险分 (25%)
45
    risk_score = (10 - min(strategy["max_drawdown"], 10)) / 10 * 100 * 0.25
46

47
    # 胜率分 (20%)
48
    winrate_score = strategy["win_rate"] * 0.20
49

50
    # Sharpe 分 (15%)
51
    sharpe_score = min(strategy["sharpe"] / 5 * 100, 100) * 0.15
52

53
    # 交易频率分 (10%)
54
    trades = strategy["trades"]
55
    if trades < 10:
56
        freq_score = 60
57
    elif 10 <= trades < 30:
58
        freq_score = 85
59
    elif 30 <= trades <= 80:
60
        freq_score = 100
61
    elif 80 < trades <= 150:
62
        freq_score = 80
63
    else:
64
        freq_score = 50
65
    freq_score *= 0.10
66

67
    total_score = profit_score + risk_score + winrate_score + sharpe_score + freq_score
68
    return round(total_score, 2)
69

70
def filter_strategies(strategies):
71
    """过滤不合格策略"""
72
    qualified = []
73
    for s in strategies:
74
        if (s["total_profit"] >= 5 and
75
            s["max_drawdown"] <= 15 and
76
            s["win_rate"] >= 50 and
77
            s["sharpe"] >= 1.0):
78
            qualified.append(s)
79
    return qualified
80

81
def rank_strategies(strategies):
82
    """策略排名"""
83
    for s in strategies:
84
        s["score"] = calculate_score(s)
85

86
    return sorted(strategies, key=lambda x: x["score"], reverse=True)
87

88
# 主程序
89
print("=" * 60)
90
print("策略自动选择工具")
91
print("=" * 60)
92

93
# 过滤
94
qualified = filter_strategies(strategies)
95
print(f"\n合格策略数: {len(qualified)} / {len(strategies)}")
96

97
# 排名
98
ranked = rank_strategies(qualified)
99

100
# 输出结果
101
print("\n策略排名（按综合评分）:")
102
print("-" * 60)
103
for i, s in enumerate(ranked, 1):
104
    print(f"{i}. {s['name']}")
105
    print(f"   总分: {s['score']} | 收益: {s['total_profit']}% | "
106
          f"回撤: {s['max_drawdown']}% | 胜率: {s['win_rate']}%")
107
    print()
108

109
# 推荐
110
print("=" * 60)
111
print("🏆 推荐策略:", ranked[0]["name"])
112
print(f"   综合评分: {ranked[0]['score']}")
113
print("=" * 60)