跳到主要內容

Feature Iteration Run #852: zz_sylvain

Status: completed | Started: 2026-07-24 08:14:59 | Ended: 2026-07-24 08:22:09

⭐ Best Accepted Iteration

Iteration 1: k_high.pct_change(12) / (1 + atr_norm) * (1 - price_positio…

baseline PF: 1.947 → modified PF: 2.401 | Cohen's d: 0.388 | p-value: 0.021

📊 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_36e183 k_high.pct_change(12) / (1 + atr_norm) … ✅ 1.947 2.401 0.9424 0.000 0.021 0.388 ✅ ACCEPT accepted
2 C8 gradient_boosting — — #1205 k_loop_iter_2_bfb3ae k_high.pct_change(12) * (1 - dist_to_hi… ✅ 1.947 1.508 -0.2477 0.000 0.579 0.155 ❌ REJECT min_p=0.579 ≥ 0.05
3 C8 gradient_boosting — — #1205 k_loop_iter_3_e77014 k_high.pct_change(12) * (1 - dist_to_hi… ✅ 1.947 1.120 -3.6347 0.000 0.086 0.542 ❌ REJECT min_p=0.086 ≥ 0.05
4 C8 gradient_boosting — — #1205 k_loop_iter_4_f0196d k_high.pct_change(12) / (1 + atr_norm) … ✅ 1.947 1.119 -0.4469 0.000 0.322 0.485 ❌ REJECT min_p=0.322 ≥ 0.05
5 C8 gradient_boosting — — #1205 k_loop_iter_5_69b6d0 k_high.pct_change(12) / (1 + atr_norm) … ✅ 1.947 1.175 -3.0769 0.000 0.079 0.481 ❌ REJECT min_p=0.079 ≥ 0.05
6 C8 gradient_boosting — — #1205 k_loop_iter_6_223860 k_high.pct_change(12) / (1 + atr_norm) … ✅ 1.947 1.175 -3.0769 0.000 0.079 0.481 ❌ REJECT min_p=0.079 ≥ 0.05
7 C8 gradient_boosting — — #1205 k_loop_iter_7_1931d4 k_momentum_pct_change * o_div_amplitude… ✅ 1.947 1.419 -0.2376 0.000 0.357 0.314 ❌ REJECT min_p=0.357 ≥ 0.05
8 C8 gradient_boosting — — #1205 k_loop_iter_8_07cfee k_momentum_pct_change * o_pct_change_12… ✅ 1.947 1.419 -0.2376 0.000 0.357 0.314 ❌ REJECT min_p=0.357 ≥ 0.05

📄 Report Markdown

# AUTOLOOP: zz_sylvain (8 iterations)

> **Phase**: Y4.7 Option B (DSL → Python code → FeatureSet candidate)
> **Target indicator**: zz_sylvain
> **Base feature_set_id**: 1205
> **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**: 429.44s
> **Total accepted**: 1
> **Total rejected**: 7

## Iterations 詳情

### Iteration 1
- DSL: `k_high.pct_change(12) / (1 + atr_norm) * (1 - price_position) * (1 - dist_to_low) * i_rsi(k_close, 14) / 100` (feature=k_loop_iter_1_36e183)
- DSL valid: True
- FeatureSet candidate id: 1205
- baseline PF: 1.947 → modified PF: 2.401
- t_stat: 0.000, p_value: 0.021, cohens_d: 0.388
- scalar_score: 0.9424
- accepted: True, reason: accepted

**Metrics summary:**
  - profit_factor: baseline=1.947, modified=2.401, diff=0.453
  - sharpe_ratio: baseline=19.162, modified=23.405, diff=4.243
  - win_rate: baseline=0.600, modified=0.548, diff=-0.052
  - max_drawdown: baseline=0.056, modified=0.056, diff=0.000
  - test_accuracy: baseline=0.992, modified=0.992, diff=-0.001

### Iteration 2
- DSL: `k_high.pct_change(12) * (1 - dist_to_high) / (1 + i_atr(k_high, k_low, k_close, 14) / k_close) * i_macd(k_close, 12, 26, 9)` (feature=k_loop_iter_2_bfb3ae)
- DSL valid: True
- FeatureSet candidate id: 1206
- baseline PF: 1.947 → modified PF: 1.508
- t_stat: 0.000, p_value: 0.579, cohens_d: 0.155
- scalar_score: -0.2477
- accepted: False, reason: min_p=0.579 ≥ 0.05

**Metrics summary:**
  - profit_factor: baseline=1.947, modified=1.508, diff=-0.440
  - sharpe_ratio: baseline=19.162, modified=10.297, diff=-8.865
  - win_rate: baseline=0.600, modified=0.448, diff=-0.152
  - max_drawdown: baseline=0.056, modified=0.061, diff=0.005
  - test_accuracy: baseline=0.992, modified=0.000, diff=-0.992

### Iteration 3
- DSL: `k_high.pct_change(12) * (1 - dist_to_high) / (1 + atr_norm) * i_rsi(k_close, 14) / 100 * (1 + hl_range)` (feature=k_loop_iter_3_e77014)
- DSL valid: True
- FeatureSet candidate id: 1207
- baseline PF: 1.947 → modified PF: 1.120
- t_stat: 0.000, p_value: 0.086, cohens_d: 0.542
- scalar_score: -3.6347
- accepted: False, reason: min_p=0.086 ≥ 0.05

**Metrics summary:**
  - profit_factor: baseline=1.947, modified=1.120, diff=-0.827
  - sharpe_ratio: baseline=19.162, modified=3.199, diff=-15.964
  - win_rate: baseline=0.600, modified=0.467, diff=-0.133
  - max_drawdown: baseline=0.056, modified=0.101, diff=0.045
  - test_accuracy: baseline=0.992, modified=0.992, diff=-0.000

### Iteration 4
- DSL: `k_high.pct_change(12) / (1 + atr_norm) * (1 - price_position) * (1 - dist_to_low) * i_rsi(k_close, 14) / 100 * (1 - k_momentum_pct_change)` (feature=k_loop_iter_4_f0196d)
- DSL valid: True
- FeatureSet candidate id: 1208
- baseline PF: 1.947 → modified PF: 1.119
- t_stat: 0.000, p_value: 0.322, cohens_d: 0.485
- scalar_score: -0.4469
- accepted: False, reason: min_p=0.322 ≥ 0.05

**Metrics summary:**
  - profit_factor: baseline=1.947, modified=1.119, diff=-0.829
  - sharpe_ratio: baseline=19.162, modified=3.554, diff=-15.609
  - win_rate: baseline=0.600, modified=0.429, diff=-0.171
  - max_drawdown: baseline=0.056, modified=0.151, diff=0.095
  - test_accuracy: baseline=0.992, modified=0.991, diff=-0.002

### Iteration 5
- DSL: `k_high.pct_change(12) / (1 + atr_norm) * (1 - price_position) * (1 - dist_to_low) * i_rsi(k_close, 14) / 100 * o_div_amplitude_12` (feature=k_loop_iter_5_69b6d0)
- DSL valid: True
- FeatureSet candidate id: 1209
- baseline PF: 1.947 → modified PF: 1.175
- t_stat: 0.000, p_value: 0.079, cohens_d: 0.481
- scalar_score: -3.0769
- accepted: False, reason: min_p=0.079 ≥ 0.05

**Metrics summary:**
  - profit_factor: baseline=1.947, modified=1.175, diff=-0.773
  - sharpe_ratio: baseline=19.162, modified=4.748, diff=-14.415
  - win_rate: baseline=0.600, modified=0.455, diff=-0.145
  - max_drawdown: baseline=0.056, modified=0.072, diff=0.016
  - test_accuracy: baseline=0.992, modified=0.992, diff=-0.000

### Iteration 6
- DSL: `k_high.pct_change(12) / (1 + atr_norm) * (1 - price_position) * (1 - dist_to_low) * i_rsi(k_close, 14) / 100 * o_pct_change_12` (feature=k_loop_iter_6_223860)
- DSL valid: True
- FeatureSet candidate id: 1210
- baseline PF: 1.947 → modified PF: 1.175
- t_stat: 0.000, p_value: 0.079, cohens_d: 0.481
- scalar_score: -3.0769
- accepted: False, reason: min_p=0.079 ≥ 0.05

**Metrics summary:**
  - profit_factor: baseline=1.947, modified=1.175, diff=-0.773
  - sharpe_ratio: baseline=19.162, modified=4.748, diff=-14.415
  - win_rate: baseline=0.600, modified=0.455, diff=-0.145
  - max_drawdown: baseline=0.056, modified=0.072, diff=0.016
  - test_accuracy: baseline=0.992, modified=0.992, diff=-0.000

### Iteration 7
- DSL: `k_momentum_pct_change * o_div_amplitude_12 / (1 + atr_norm) * i_macd(k_close, 12, 26, 9)` (feature=k_loop_iter_7_1931d4)
- DSL valid: True
- FeatureSet candidate id: 1211
- baseline PF: 1.947 → modified PF: 1.419
- t_stat: 0.000, p_value: 0.357, cohens_d: 0.314
- scalar_score: -0.2376
- accepted: False, reason: min_p=0.357 ≥ 0.05

**Metrics summary:**
  - profit_factor: baseline=1.947, modified=1.419, diff=-0.529
  - sharpe_ratio: baseline=19.162, modified=10.520, diff=-8.643
  - win_rate: baseline=0.600, modified=0.483, diff=-0.117
  - max_drawdown: baseline=0.056, modified=0.252, diff=0.196
  - test_accuracy: baseline=0.992, modified=0.981, diff=-0.012

### Iteration 8
- DSL: `k_momentum_pct_change * o_pct_change_12 / (1 + atr_norm) * i_macd(k_close, 12, 26, 9) * (1 - dist_to_high)` (feature=k_loop_iter_8_07cfee)
- DSL valid: True
- FeatureSet candidate id: 1212
- baseline PF: 1.947 → modified PF: 1.419
- t_stat: 0.000, p_value: 0.357, cohens_d: 0.314
- scalar_score: -0.2376
- accepted: False, reason: min_p=0.357 ≥ 0.05

**Metrics summary:**
  - profit_factor: baseline=1.947, modified=1.419, diff=-0.529
  - sharpe_ratio: baseline=19.162, modified=10.520, diff=-8.643
  - win_rate: baseline=0.600, modified=0.483, diff=-0.117
  - max_drawdown: baseline=0.056, modified=0.252, diff=0.196
  - test_accuracy: baseline=0.992, modified=0.981, diff=-0.012

📊 Multi-Objective Metrics

Iteration 1: k_loop_iter_1_36e183

Y Label: C8 | Parent FS: #7

Objective mode: scalarize

Scalar score: 0.9424

MetricBaselineModifiedDiff
win_rate 0.600 0.548 -0.052
max_drawdown 0.056 0.056 0.000
sharpe_ratio 19.162 23.405 4.243
profit_factor 1.947 2.401 0.453
test_accuracy 0.992 0.992 -0.001

Iteration 2: k_loop_iter_2_bfb3ae

Y Label: C8 | Parent FS: #1205

Objective mode: scalarize

Scalar score: -0.2477

MetricBaselineModifiedDiff
win_rate 0.600 0.448 -0.152
max_drawdown 0.056 0.061 0.005
sharpe_ratio 19.162 10.297 -8.865
profit_factor 1.947 1.508 -0.440
test_accuracy 0.992 0.000 -0.992

Iteration 3: k_loop_iter_3_e77014

Y Label: C8 | Parent FS: #1205

Objective mode: scalarize

Scalar score: -3.6347

MetricBaselineModifiedDiff
win_rate 0.600 0.467 -0.133
max_drawdown 0.056 0.101 0.045
sharpe_ratio 19.162 3.199 -15.964
profit_factor 1.947 1.120 -0.827
test_accuracy 0.992 0.992 0.000

Iteration 4: k_loop_iter_4_f0196d

Y Label: C8 | Parent FS: #1205

Objective mode: scalarize

Scalar score: -0.4469

MetricBaselineModifiedDiff
win_rate 0.600 0.429 -0.171
max_drawdown 0.056 0.151 0.095
sharpe_ratio 19.162 3.554 -15.609
profit_factor 1.947 1.119 -0.829
test_accuracy 0.992 0.991 -0.002

Iteration 5: k_loop_iter_5_69b6d0

Y Label: C8 | Parent FS: #1205

Objective mode: scalarize

Scalar score: -3.0769

MetricBaselineModifiedDiff
win_rate 0.600 0.455 -0.145
max_drawdown 0.056 0.072 0.016
sharpe_ratio 19.162 4.748 -14.415
profit_factor 1.947 1.175 -0.773
test_accuracy 0.992 0.992 0.000

Iteration 6: k_loop_iter_6_223860

Y Label: C8 | Parent FS: #1205

Objective mode: scalarize

Scalar score: -3.0769

MetricBaselineModifiedDiff
win_rate 0.600 0.455 -0.145
max_drawdown 0.056 0.072 0.016
sharpe_ratio 19.162 4.748 -14.415
profit_factor 1.947 1.175 -0.773
test_accuracy 0.992 0.992 0.000

Iteration 7: k_loop_iter_7_1931d4

Y Label: C8 | Parent FS: #1205

Objective mode: scalarize

Scalar score: -0.2376

MetricBaselineModifiedDiff
win_rate 0.600 0.483 -0.117
max_drawdown 0.056 0.252 0.196
sharpe_ratio 19.162 10.520 -8.643
profit_factor 1.947 1.419 -0.529
test_accuracy 0.992 0.981 -0.012

Iteration 8: k_loop_iter_8_07cfee

Y Label: C8 | Parent FS: #1205

Objective mode: scalarize

Scalar score: -0.2376

MetricBaselineModifiedDiff
win_rate 0.600 0.483 -0.117
max_drawdown 0.056 0.252 0.196
sharpe_ratio 19.162 10.520 -8.643
profit_factor 1.947 1.419 -0.529
test_accuracy 0.992 0.981 -0.012

📁 Offline Report

Markdown backup: reports/autoloop/zz_sylvain/20260724_162209.md