fix: address pre-push review findings

backtesting/engine.py

@@ -24,6 +24,7 @@ ML_WEIGHTS_PATH = os.path.join(BASE_DIR, "config", "ml_weights.json")
 
 _BACKTEST_CACHE = {}
 _BACKTEST_CACHE_LOCK = threading.Lock()
+_BACKTEST_CACHE_LIMIT = 4
 
 # Score brackets matching the dashboard assessment levels
 BRACKETS = SCORE_BRACKETS
@@ -422,7 +423,12 @@ def clear_backtest_cache():
 
 def _add_return_statistics(stats, period, returns):
     """Add return summaries and a moving-block-bootstrap mean interval."""
-    summary = summarize_returns(returns, block_size=min(30, len(returns)), n_resamples=400)
+    horizon_days = int(period.removesuffix("d"))
+    summary = summarize_returns(
+        returns,
+        block_size=min(horizon_days, len(returns)),
+        n_resamples=400,
+    )
     stats[f"avg_{period}"] = summary["mean"]
     stats[f"median_{period}"] = summary["median"]
     stats[f"win_rate_{period}"] = summary["win_rate"]
@@ -449,8 +455,9 @@ def run_backtest(ml_mode=False):
 
     result = _compute_backtest(ml_mode=ml_mode)
     with _BACKTEST_CACHE_LOCK:
-        _BACKTEST_CACHE.clear()
         _BACKTEST_CACHE[signature] = copy.deepcopy(result)
+        while len(_BACKTEST_CACHE) > _BACKTEST_CACHE_LIMIT:
+            _BACKTEST_CACHE.pop(next(iter(_BACKTEST_CACHE)))
     return copy.deepcopy(result)
 
 
@@ -720,7 +727,7 @@ def _compute_backtest(ml_mode=False):
             # Include per-metric values (raw metric value, not score)
             metric_vals = d.get("metric_values", {})
             if metric_vals:
-                entry["metrics"] = metric_vals
+                entry["metric_values"] = metric_vals
             chart_data.append(entry)
 
     if not ml_mode: