跳到主要內容

Feature Iteration Run #358: zz_sylvain

Status: completed | Started: 2026-07-16 15:09:01 | Ended: 2026-07-16 15:14:59

⭐ Best Accepted Iteration

Iteration 1: k_high.pct_change(12)

baseline PF: 1.037 → modified PF: 1.062 | Cohen's d: 0.000 | p-value: 1.000

📊 Iterations

Iter Y Label Model Warm Proxy Parent FS Feature DSL DSL Valid Baseline PF Modified PF Scalar Score t-stat p-value Cohen's d Result Reason
1 — gradient_boosting — — — K_loop_iter_1_563283 k_high.pct_change(12) ✅ 1.037 1.062 0.1812 0.000 1.000 0.000 ✅ ACCEPT accepted
2 — gradient_boosting — — — K_loop_iter_1_563283 k_high.pct_change(12) ✅ 1.198 1.189 -0.0605 0.000 1.000 0.000 ❌ REJECT p=1.000 ≥ 0.05, cohens_d=0.000 ≤ 0.3
3 — gradient_boosting — — — K_loop_iter_1_563283 k_high.pct_change(12) ✅ 1.219 1.255 0.2410 0.000 1.000 0.000 ✅ ACCEPT accepted
4 — gradient_boosting — — — K_loop_iter_1_563283 k_high.pct_change(12) ✅ 0.967 0.919 -0.4836 0.000 1.000 0.000 ❌ REJECT p=1.000 ≥ 0.05, cohens_d=0.000 ≤ 0.3, constraints: ['sharpe_ratio >= 1.0: got -3.1653']
5 — gradient_boosting — — — K_loop_iter_1_563283 k_high.pct_change(12) ✅ 1.005 1.099 0.8877 0.000 1.000 0.000 ✅ ACCEPT accepted

📄 Report Markdown

# AUTOLOOP: zz_sylvain (5 iterations)

> **Phase**: Y4.7 Option B (DSL → Python code → FeatureSet candidate)
> **Target indicator**: zz_sylvain
> **Base feature_set_id**: 7
> **Claim**: (default) Adding k_high_pct_change_12 improves Sharpe by 10%
> **Objective mode**: scalarize
> **Metrics**: ['profit_factor', 'sharpe_ratio', 'win_rate', 'max_drawdown', 'test_accuracy']
> **Total duration**: 357.59s
> **Total accepted**: 3
> **Total rejected**: 2

## Iterations 詳情

### Iteration 1
- DSL: `k_high.pct_change(12)` (feature=K_loop_iter_1_563283)
- DSL valid: True
- FeatureSet candidate id: 328
- baseline PF: 1.037 → modified PF: 1.062
- t_stat: 0.000, p_value: 1.000, cohens_d: 0.000
- scalar_score: 0.1812
- accepted: True, reason: accepted

**Metrics summary:**
  - profit_factor: baseline=1.037, modified=1.062, diff=0.026
  - sharpe_ratio: baseline=0.995, modified=1.675, diff=0.680
  - win_rate: baseline=0.419, modified=0.438, diff=0.019
  - max_drawdown: baseline=0.120, modified=0.118, diff=-0.003
  - test_accuracy: baseline=0.971, modified=0.970, diff=-0.000

### Iteration 2
- DSL: `k_high.pct_change(12)` (feature=K_loop_iter_1_563283)
- DSL valid: True
- FeatureSet candidate id: 329
- baseline PF: 1.198 → modified PF: 1.189
- t_stat: 0.000, p_value: 1.000, cohens_d: 0.000
- scalar_score: -0.0605
- accepted: False, reason: p=1.000 ≥ 0.05, cohens_d=0.000 ≤ 0.3

**Metrics summary:**
  - profit_factor: baseline=1.198, modified=1.189, diff=-0.009
  - sharpe_ratio: baseline=5.537, modified=5.294, diff=-0.243
  - win_rate: baseline=0.443, modified=0.432, diff=-0.011
  - max_drawdown: baseline=0.082, modified=0.117, diff=0.034
  - test_accuracy: baseline=0.973, modified=0.974, diff=0.001

### Iteration 3
- DSL: `k_high.pct_change(12)` (feature=K_loop_iter_1_563283)
- DSL valid: True
- FeatureSet candidate id: 330
- baseline PF: 1.219 → modified PF: 1.255
- t_stat: 0.000, p_value: 1.000, cohens_d: 0.000
- scalar_score: 0.2410
- accepted: True, reason: accepted

**Metrics summary:**
  - profit_factor: baseline=1.219, modified=1.255, diff=0.036
  - sharpe_ratio: baseline=5.847, modified=6.783, diff=0.936
  - win_rate: baseline=0.461, modified=0.470, diff=0.009
  - max_drawdown: baseline=0.187, modified=0.151, diff=-0.037
  - test_accuracy: baseline=0.970, modified=0.969, diff=-0.001

### Iteration 4
- DSL: `k_high.pct_change(12)` (feature=K_loop_iter_1_563283)
- DSL valid: True
- FeatureSet candidate id: 331
- baseline PF: 0.967 → modified PF: 0.919
- t_stat: 0.000, p_value: 1.000, cohens_d: 0.000
- scalar_score: -0.4836
- accepted: False, reason: p=1.000 ≥ 0.05, cohens_d=0.000 ≤ 0.3, constraints: ['sharpe_ratio >= 1.0: got -3.1653']

**Metrics summary:**
  - profit_factor: baseline=0.967, modified=0.919, diff=-0.048
  - sharpe_ratio: baseline=-1.275, modified=-3.165, diff=-1.891
  - win_rate: baseline=0.565, modified=0.552, diff=-0.013
  - max_drawdown: baseline=0.160, modified=0.199, diff=0.039
  - test_accuracy: baseline=0.976, modified=0.976, diff=0.001

### Iteration 5
- DSL: `k_high.pct_change(12)` (feature=K_loop_iter_1_563283)
- DSL valid: True
- FeatureSet candidate id: 332
- baseline PF: 1.005 → modified PF: 1.099
- t_stat: 0.000, p_value: 1.000, cohens_d: 0.000
- scalar_score: 0.8877
- accepted: True, reason: accepted

**Metrics summary:**
  - profit_factor: baseline=1.005, modified=1.099, diff=0.094
  - sharpe_ratio: baseline=0.176, modified=3.626, diff=3.450
  - win_rate: baseline=0.497, modified=0.500, diff=0.003
  - max_drawdown: baseline=0.086, modified=0.062, diff=-0.024
  - test_accuracy: baseline=0.982, modified=0.982, diff=-0.001

📊 Multi-Objective Metrics

Iteration 1: K_loop_iter_1_563283

Y Label: — | Parent FS: #—

Objective mode: scalarize

Scalar score: 0.1812

MetricBaselineModifiedDiff
win_rate 0.419 0.438 0.019
max_drawdown 0.120 0.118 -0.003
sharpe_ratio 0.995 1.675 0.680
profit_factor 1.037 1.062 0.026
test_accuracy 0.971 0.970 0.000

Iteration 2: K_loop_iter_1_563283

Y Label: — | Parent FS: #—

Objective mode: scalarize

Scalar score: -0.0605

MetricBaselineModifiedDiff
win_rate 0.443 0.432 -0.011
max_drawdown 0.082 0.117 0.034
sharpe_ratio 5.537 5.294 -0.243
profit_factor 1.198 1.189 -0.009
test_accuracy 0.973 0.974 0.001

Iteration 3: K_loop_iter_1_563283

Y Label: — | Parent FS: #—

Objective mode: scalarize

Scalar score: 0.2410

MetricBaselineModifiedDiff
win_rate 0.461 0.470 0.009
max_drawdown 0.187 0.151 -0.037
sharpe_ratio 5.847 6.783 0.936
profit_factor 1.219 1.255 0.036
test_accuracy 0.970 0.969 -0.001

Iteration 4: K_loop_iter_1_563283

Y Label: — | Parent FS: #—

Objective mode: scalarize

Scalar score: -0.4836

Failed constraints: sharpe_ratio >= 1.0: got -3.1653

MetricBaselineModifiedDiff
win_rate 0.565 0.552 -0.013
max_drawdown 0.160 0.199 0.039
sharpe_ratio -1.275 -3.165 -1.891
profit_factor 0.967 0.919 -0.048
test_accuracy 0.976 0.976 0.001

Iteration 5: K_loop_iter_1_563283

Y Label: — | Parent FS: #—

Objective mode: scalarize

Scalar score: 0.8877

MetricBaselineModifiedDiff
win_rate 0.497 0.500 0.003
max_drawdown 0.086 0.062 -0.024
sharpe_ratio 0.176 3.626 3.450
profit_factor 1.005 1.099 0.094
test_accuracy 0.982 0.982 -0.001

📁 Offline Report

Markdown backup: reports/autoloop/zz_sylvain/20260716_231459.md