From: Jérôme Benoit <jerome.benoit@piment-noir.org>
Date: Wed, 18 Feb 2026 23:07:20 +0000 (+0100)
Subject: refactor(zigzag): replace ad-hoc x/(x+y) normalization with pure ratios + MinMaxScaler
X-Git-Url: https://git.piment-noir.org/?a=commitdiff_plain;h=b85991862c288fc23c6e36600fc62a5c2718c7c1;p=freqai-strategies.git

refactor(zigzag): replace ad-hoc x/(x+y) normalization with pure ratios + MinMaxScaler

Ultraworked with [Sisyphus](https://github.com/code-yeongyu/oh-my-opencode)

Co-authored-by: Sisyphus <clio-agent@sisyphuslabs.ai>
---

diff --git a/quickadapter/user_data/strategies/Utils.py b/quickadapter/user_data/strategies/Utils.py
index 128d974..47d950e 100644
--- a/quickadapter/user_data/strategies/Utils.py
+++ b/quickadapter/user_data/strategies/Utils.py
@@ -1732,6 +1732,27 @@ def zigzag(
         candidate_pivot_pos = -1
         candidate_pivot_value_log = np.nan
 
+    def minmax_scale(values: list[float]) -> list[float]:
+        """Scale values to [0, 1] range preserving NaN.
+
+        Edge cases:
+        - Empty: return []
+        - All NaN: return all NaN
+        - All same finite value: return [0.5, 0.5, ...]
+        - Normal: return MinMaxScaler result as list
+        """
+        if not values:
+            return values
+        arr = np.asarray(values, dtype=float)
+        valid_mask = np.isfinite(arr)
+        if not valid_mask.any():
+            return values  # all NaN → preserve
+        lo, hi = np.nanmin(arr), np.nanmax(arr)
+        if np.isclose(lo, hi):
+            return [0.5 if np.isfinite(v) else np.nan for v in values]
+        scaled = (arr - lo) / (hi - lo)
+        return scaled.tolist()
+
     def calculate_pivot_metrics(
         *,
         previous_pos: int,
@@ -1756,7 +1777,7 @@ def zigzag(
         median_threshold_log = np.nanmedian(np.log1p(thresholds[start_pos:end_pos]))
 
         amplitude_threshold_ratio = (
-            amplitude / (amplitude + median_threshold_log)
+            amplitude / median_threshold_log
             if np.isfinite(median_threshold_log) and median_threshold_log > 0
             else np.nan
         )
@@ -1816,7 +1837,7 @@ def zigzag(
             and np.isfinite(median_volume)
             and median_volume > 0
         ):
-            return avg_volume_per_candle / (avg_volume_per_candle + median_volume)
+            return avg_volume_per_candle / median_volume
         return np.nan
 
     def calculate_pivot_efficiency_ratio(
@@ -2094,10 +2115,10 @@ def zigzag(
         pivots_indices,
         pivots_values_log,
         pivots_directions,
-        pivots_amplitudes,
-        pivots_amplitude_threshold_ratios,
-        pivots_volume_rates,
-        pivots_speeds,
+        minmax_scale(pivots_amplitudes),
+        minmax_scale(pivots_amplitude_threshold_ratios),
+        minmax_scale(pivots_volume_rates),
+        minmax_scale(pivots_speeds),
         pivots_efficiency_ratios,
         pivots_volume_weighted_efficiency_ratios,
     )