From 2ff8ecaf528c512a3a0fa0411857a11ae0b307fe Mon Sep 17 00:00:00 2001
From: claude_dev <claude-dev@sanguo.local>
Date: Sun, 2 Aug 2026 20:54:50 +0800
Subject: [PATCH] =?UTF-8?q?perf(metrics):=20rolling=20alpha/beta=20?=
 =?UTF-8?q?=E5=90=91=E9=87=8F=E5=8C=96(O(n=C2=B2)=E2=86=92O(n),=E6=B2=BB?=
 =?UTF-8?q?=E5=88=86=E9=92=9F=E7=BA=A7=E5=8D=A1=E6=AD=BB)?=
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

compute_metrics rolling alpha/beta 原 for 循环每点 iloc[:i+1].cov/var = O(n²)。日线 111 天没事,分钟级 daily_df(万行)卡死(cta_c287f4b5 卡 5min CPU88%)。改 pandas expanding 向量化(cov/var/mean O(n)),数值与原循环一致(beta/alpha diff<1e-15,NaN 位置匹配)。pytest 6 绿。触发:b2c41c7 降级让 compute_metrics 首次在 benchmark 有数据时真跑(此前 benchmark 空被 skip),暴露 rolling O(n²) 性能 bug。
---
 sanguo_backtest/metrics.py | 16 +++++++---------
 1 file changed, 7 insertions(+), 9 deletions(-)

diff --git a/sanguo_backtest/metrics.py b/sanguo_backtest/metrics.py
index c3a54d4..cd26a3b 100644
--- a/sanguo_backtest/metrics.py
+++ b/sanguo_backtest/metrics.py
@@ -69,15 +69,13 @@ def compute_metrics(
     equity = empyrical.cum_returns(s)
     bench_curve = empyrical.cum_returns(b)
     # rolling alpha/beta (expanding window 用于画图，口径由 scalars 保证)
-    roll_beta = pd.Series(index=s.index, dtype=float)
-    roll_alpha = pd.Series(index=s.index, dtype=float)
-    for i in range(len(s)):
-        sub = aligned.iloc[: i + 1]
-        if len(sub) >= 2 and sub["b"].var() > 0:
-            beta = sub["s"].cov(sub["b"]) / sub["b"].var()
-            alpha = sub["s"].mean() - beta * sub["b"].mean()
-            roll_beta.iloc[i] = beta
-            roll_alpha.iloc[i] = alpha * period
+    # 向量化 expanding (O(n)) 替代原 for 循环 (O(n²))——分钟级 daily_df(万行)原循环卡死
+    # (cta_c287f4b5 卡 5min CPU88% 根因)。expanding.cov/.var/.mean 数值与原循环一致
+    # (ddof=1)；var_b<=0(含单点 NaN)处 beta=NaN，与原 len>=2 且 var>0 守卫等价。
+    cov_sb = aligned["s"].expanding().cov(aligned["b"])
+    var_b = aligned["b"].expanding().var()
+    roll_beta = cov_sb / var_b.where(var_b > 0)
+    roll_alpha = (aligned["s"].expanding().mean() - roll_beta * aligned["b"].expanding().mean()) * period
     # drawdown: 从峰值回落 (值 <= 0)
     cummax = equity.cummax()
     drawdown = (equity - cummax) / cummax