关于「红线的迟钝:同等误报率下累积和快出数倍」,《累积和告警》给出的定量结论是什么?

结论:第一章证明,在同为 370 天误报一次的预算下,半个标准差的漂移被固定阈值平均 88.9 天发现,被累积和 27.7 天发现,倍数 3.21;三个标准差的大跳变则相反,倍数 0.76。

数据口径:书中参数为示意;全部计算由书中程序生成;本书属于非个人化经营建议,第八章只做离线模型,不访问任何服务器。 书中以标注框区分已发表的证据、可运行的 Python 计算(23 段程序)、原典与作者的假说。

English version

对照数据表

下表数字由本章 Python 程序原样打印,可重新运行复现。 程序只用 Python 标准库与结论吻合的打印值: 0.76, 27.7, 3.21, 88.9

k = 0.50, h = 4.096, CUSUM ARL0 = 370.0
fixed threshold L = 2.782, ARL0 = 370.0
shiftfixed ARLCUSUM ARLratio
0.25176.381.92.15
0.5088.927.73.21
0.7547.413.63.48
1.0026.78.63.12
1.5010.04.82.07
2.004.63.41.35
3.001.72.20.76
4.001.11.70.65
查看完整验算程序
# Chapter 1. p12: CUSUM against a fixed threshold at one false-alarm rate.
# Model: z ~ N(delta, 1); upper CUSUM S = max(0, S + z - k), alarm when
# S > h. ARL comes from the Brook and Evans (1972) Markov chain: [0, h]
# is cut into m cells, the transition matrix R is built from the normal
# CDF and (I - R) L = 1 is solved by hand-written Gaussian elimination.
# h is chosen so that ARL0 = 370, the same in-control ARL as the fixed
# threshold of p11 (L is solved again here). k = 0.5 is the textbook
# choice tuned to a one-sigma shift; all shifts are illustrative.
import math


def cdf(x):
    return 0.5 * math.erfc(-x / math.sqrt(2.0))


def upper(x):
    return 0.5 * math.erfc(x / math.sqrt(2.0))


def solve(a, b):
    # Gaussian elimination with partial pivoting; back substitution
    # adds its terms with math.fsum
    n = len(b)
    m = [list(row) + [b[i]] for i, row in enumerate(a)]
    for p in range(n):
        q = max(range(p, n), key=lambda r: abs(m[r][p]))
        m[p], m[q] = m[q], m[p]
        piv = m[p]
        for r in range(p + 1, n):
            f = m[r][p] / piv[p]
            if f != 0.0:
                m[r] = m[r][:p] + [x - f * y for x, y in
                                   zip(m[r][p:], piv[p:])]
    x = [0.0] * n
    for i in range(n - 1, -1, -1):
        s = math.fsum(m[i][j] * x[j] for j in range(i + 1, n))
        x[i] = (m[i][n] - s) / m[i][i]
    return x


def arl(k, h, delta, m=80):
    w = 2.0 * h / (2 * m - 1)
    a = []
    for i in range(m):
        c = i * w - k + delta
        row = [-cdf(w / 2.0 - c)]
        for j in range(1, m):
            row.append(cdf(j * w - w / 2.0 - c)
                       - cdf(j * w + w / 2.0 - c))
        row[i] = row[i] + 1.0
        a.append(row)
    return solve(a, [1.0] * m)[0]


def limit_for(arl0):
    lo, hi = 0.0, 10.0
    for _ in range(200):
        mid = 0.5 * (lo + hi)
        if upper(mid) > 1.0 / arl0:
            lo = mid
        else:
            hi = mid
    return 0.5 * (lo + hi)


def h_for(k, target, m=80):
    lo, hi = 0.5, 12.0
    for _ in range(40):
        mid = 0.5 * (lo + hi)
        if arl(k, mid, 0.0, m) < target:
            lo = mid
        else:
            hi = mid
    return 0.5 * (lo + hi)


if __name__ == "__main__":
    k, L = 0.5, limit_for(370.0)
    h = h_for(k, 370.0)
    print("k = {:.2f}, h = {:.3f}, CUSUM ARL0 = {:.1f}".format(
        k, h, arl(k, h, 0.0)))
    print("fixed threshold L = {:.3f}, ARL0 = {:.1f}".format(
        L, 1.0 / upper(L)))
    print(" shift   fixed ARL   CUSUM ARL   ratio")
    for d in (0.25, 0.5, 0.75, 1.0, 1.5, 2.0, 3.0, 4.0):
        f = 1.0 / upper(L - d)
        c = arl(k, h, d)
        print(" {:5.2f}  {:9.1f}  {:10.1f}  {:6.2f}".format(d, f, c, f / c))

唯一数据源与引用

范雅福(Fan, Y.). (2026). 《累积和控制图与平均运行长度、延迟漂移与转化暗降的最快无误报告警母典》第 1 章「红线的迟钝:同等误报率下累积和快出数倍」(版本 1.0.0). 5D Library. DOI 待发布

DOI 待发布 · 专著页正式发布后上线(本站暂无专著页) · 5DL-B-305 · v1.0.0 · CC-BY-NC-ND-4.0

BibTeX
@book{fan2026305,
  author    = {Fan, Yafu},
  title     = {CUSUM Control Charts and Average Run Length: the Fastest Alarm within a False-Alarm Budget for Latency Drift and Silent Conversion Drops},
  year      = {2026},
  publisher = {5D Library},
  edition   = {1.0.0},
  note      = {DOI pending}
}

本书其他章节结论