跳到主要內容

Feature Iteration Run #851: zz_sylvain

Status: completed | Started: 2026-07-24 07:59:38 | Ended: 2026-07-24 08:02:12

⭐ Best Accepted Iteration

Iteration 1: k_high.pct_change(12) * (1 + hl_range) * (1 - dist_to_high)…

baseline PF: 1.947 → modified PF: 2.514 | Cohen's d: 0.577 | p-value: 0.000

📊 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 C8 gradient_boosting — — #7 k_loop_iter_1_d262e5 k_high.pct_change(12) * (1 + hl_range) … ✅ 1.947 2.514 1.7780 0.000 0.000 0.577 ✅ ACCEPT accepted
2 C8 gradient_boosting — — #1202 k_loop_iter_2_0bd1a0 k_high.pct_change(12) * (1 + hl_range) … ✅ 1.947 1.301 -0.3396 0.000 0.586 0.317 ❌ REJECT min_p=0.586 ≥ 0.05
3 C8 gradient_boosting — — #1202 k_loop_iter_3_ca04cb k_high.pct_change(12) * (1 + hl_range) … ✅ 1.947 1.258 -0.3277 0.000 0.466 0.320 ❌ REJECT min_p=0.466 ≥ 0.05

📄 Report Markdown

# AUTOLOOP: zz_sylvain (3 iterations)

> **Phase**: Y4.7 Option B (DSL → Python code → FeatureSet candidate)
> **Target indicator**: zz_sylvain
> **Base feature_set_id**: 1202
> **Claim**: (default) Propose a new k_* feature that improves risk-adjusted returns
> **Objective mode**: scalarize
> **Metrics**: ['profit_factor', 'sharpe_ratio', 'win_rate', 'max_drawdown', 'test_accuracy']
> **Total duration**: 153.75s
> **Total accepted**: 1
> **Total rejected**: 2

## Iterations 詳情

### Iteration 1
- DSL: `k_high.pct_change(12) * (1 + hl_range) * (1 - dist_to_high) * (1 + i_macd(k_close, 12, 26, 9) / (1 + atr_norm))` (feature=k_loop_iter_1_d262e5)
- DSL valid: True
- FeatureSet candidate id: 1202
- baseline PF: 1.947 → modified PF: 2.514
- t_stat: 0.000, p_value: 0.000, cohens_d: 0.577
- scalar_score: 1.7780
- accepted: True, reason: accepted

**Metrics summary:**
  - profit_factor: baseline=1.947, modified=2.514, diff=0.566
  - sharpe_ratio: baseline=19.162, modified=25.564, diff=6.401
  - win_rate: baseline=0.600, modified=0.643, diff=0.043
  - max_drawdown: baseline=0.056, modified=0.042, diff=-0.014
  - test_accuracy: baseline=0.992, modified=0.993, diff=0.000

### Iteration 2
- DSL: `k_high.pct_change(12) * (1 + hl_range) * (1 - dist_to_high) * (1 + i_rsi(k_close, 14) / (1 + atr_norm))` (feature=k_loop_iter_2_0bd1a0)
- DSL valid: True
- FeatureSet candidate id: 1203
- baseline PF: 1.947 → modified PF: 1.301
- t_stat: 0.000, p_value: 0.586, cohens_d: 0.317
- scalar_score: -0.3396
- accepted: False, reason: min_p=0.586 ≥ 0.05

**Metrics summary:**
  - profit_factor: baseline=1.947, modified=1.301, diff=-0.646
  - sharpe_ratio: baseline=19.162, modified=6.894, diff=-12.268
  - win_rate: baseline=0.600, modified=0.425, diff=-0.175
  - max_drawdown: baseline=0.056, modified=0.053, diff=-0.003
  - test_accuracy: baseline=0.992, modified=0.000, diff=-0.992

### Iteration 3
- DSL: `k_high.pct_change(12) * (1 + hl_range) * (1 - dist_to_high) * (1 - dist_to_low) / (1 + atr_norm)` (feature=k_loop_iter_3_ca04cb)
- DSL valid: True
- FeatureSet candidate id: 1204
- baseline PF: 1.947 → modified PF: 1.258
- t_stat: 0.000, p_value: 0.466, cohens_d: 0.320
- scalar_score: -0.3277
- accepted: False, reason: min_p=0.466 ≥ 0.05

**Metrics summary:**
  - profit_factor: baseline=1.947, modified=1.258, diff=-0.690
  - sharpe_ratio: baseline=19.162, modified=7.336, diff=-11.827
  - win_rate: baseline=0.600, modified=0.509, diff=-0.091
  - max_drawdown: baseline=0.056, modified=0.305, diff=0.249
  - test_accuracy: baseline=0.992, modified=0.974, diff=-0.019

📊 Multi-Objective Metrics

Iteration 1: k_loop_iter_1_d262e5

Y Label: C8 | Parent FS: #7

Objective mode: scalarize

Scalar score: 1.7780

MetricBaselineModifiedDiff
win_rate 0.600 0.643 0.043
max_drawdown 0.056 0.042 -0.014
sharpe_ratio 19.162 25.564 6.401
profit_factor 1.947 2.514 0.566
test_accuracy 0.992 0.993 0.000

Iteration 2: k_loop_iter_2_0bd1a0

Y Label: C8 | Parent FS: #1202

Objective mode: scalarize

Scalar score: -0.3396

MetricBaselineModifiedDiff
win_rate 0.600 0.425 -0.175
max_drawdown 0.056 0.053 -0.003
sharpe_ratio 19.162 6.894 -12.268
profit_factor 1.947 1.301 -0.646
test_accuracy 0.992 0.000 -0.992

Iteration 3: k_loop_iter_3_ca04cb

Y Label: C8 | Parent FS: #1202

Objective mode: scalarize

Scalar score: -0.3277

MetricBaselineModifiedDiff
win_rate 0.600 0.509 -0.091
max_drawdown 0.056 0.305 0.249
sharpe_ratio 19.162 7.336 -11.827
profit_factor 1.947 1.258 -0.690
test_accuracy 0.992 0.974 -0.019

📁 Offline Report

Markdown backup: reports/autoloop/zz_sylvain/20260724_160212.md