Feature Iteration Run #1875: zz_sylvain

# AUTOLOOP: zz_sylvain (3 iterations)

> **Phase**: Y4.7 Option B (DSL → Python code → FeatureSet candidate)
> **Target indicator**: zz_sylvain
> **Base feature_set_id**: 1748
> **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**: 3285.64s
> **Total accepted**: 0
> **Total rejected**: 3

## Iterations 詳情

### Iteration 1
- DSL: `macd_line = ta.ema(close, 12) - ta.ema(close, 26)
signal = ta.ema(macd_line, 9)
hist = macd_line - signal
atr_v = ta.atr(high, low, close, 14)
vol_sma = ta.sma(volume, 20)
vol_r = volume / (vol_sma + 1e-8)
mom_fast = (close - close[10]) / (close[10])
result = (hist / (close * (1 + atr_v / close))) * vol_r * (1 + mom_fast)` (feature=k_iter1_macd_hist_atr_vol_mom)
- DSL valid: False
- FeatureSet candidate id: None
- baseline PF: 0.000 → modified PF: 0.000
- t_stat: 0.000, p_value: 1.000, cohens_d: 0.000
- scalar_score: 0.0000
- accepted: False, reason: translator_verify: 3 輪修復仍未通過 — max_diff=inf (tol=0.001), nan_mismatch=999, error=OO exec error: 'KMacdHistAtrVolMom' object has no attribute 'ema'

### Iteration 2
- DSL: `fast = ta.ema(close, 8)
slow = ta.ema(close, 34)
macd_line = fast - slow
atr_v = ta.atr(high, low, close, 14)
denom = close * (1 + atr_v / close)
up_state = close > slow ? 1 : -1
result = (macd_line / denom) * up_state` (feature=k_iter2_macd_line_atr_trend_gate)
- DSL valid: False
- FeatureSet candidate id: None
- baseline PF: 0.000 → modified PF: 0.000
- t_stat: 0.000, p_value: 1.000, cohens_d: 0.000
- scalar_score: 0.0000
- accepted: False, reason: translator_verify: 3 輪修復仍未通過 — max_diff=inf (tol=0.001), nan_mismatch=999, error=OO exec error: 'KMacdLineAtrTrendGate' object has no attribute 'ema'

### Iteration 3
- DSL: `result = ta.rsi(close, 14) / 100 - 0.5` (feature=k_iter3_rsi_centered)
- DSL valid: True
- FeatureSet candidate id: 1842
- baseline PF: 1.836 → modified PF: 1.542
- t_stat: 0.000, p_value: 0.150, cohens_d: 0.227
- scalar_score: -1.6577
- accepted: False, reason: normalized_score -0.1232 <= 0.0, profit_factor_delta -0.2941 < 0.25, max_drawdown_delta -0.0235 < -0.02, sharpe_delta -8.0317 (-38.52%) < -10%, hypothesis test not confirmed, DSR gate failed: champion SR=3.1268 ≤ expected_max=3.3474 (N=29 effective trials, PSR=0.4302)

**Metrics summary:**
  - profit_factor: baseline=1.836, modified=1.542, diff=-0.294
  - sharpe_ratio: baseline=20.850, modified=12.818, diff=-8.032
  - win_rate: baseline=0.566, modified=0.538, diff=-0.028
  - max_drawdown: baseline=0.129, modified=0.105, diff=-0.023
  - test_accuracy: baseline=0.992, modified=0.972, diff=-0.020