关于「实证闭环:剪枝器接入、月度归约与预注册检验」,《冲突多半是冗余》给出的定量结论是什么?

结论:第八章给出剪枝器接入的八步月度流程、蒸馏小工具「规则冲突归约器」与预注册检验:60 对盲判配对、拒绝界 23、检验水平 0.0462。书中以标注框区分已发表的证据、可运行的 Python 计算(19 段程序)、原典与作者的假说。

数据口径:书中规则库与数值均为示意;全部计算由书中程序生成;本书属非个人化经营方法,不构成法律、合规或经营决策意见,规则的增删须由主事者人工放行。 第一章正面陈述并量化命题:一千条示意规则里,标红配对中真矛盾只占 0.0392,裁决后 971 条是冗余,只剩 18 条约束性规则;越界规则占一成时真矛盾比例也只有 0.2667,但无解指标已遍及全部。

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: 0.0462, 23, 60

H0 share >= 0.5, one-sided exact binomial, alpha 0.05
true sharen(80%)reject if <=n(90%)reject if <=
0.1081112
0.20185237
0.3037135320
0.3569279338
plan n = 60: reject H0 if at most 23 are true contradictions
size under 0.5 = 0.0462; power at 0.20 = 0.9997, at 0.35 = 0.7529
查看完整验算程序
# Chapter 8. p82: how many flagged conflicts must an audit label?
# Model: an auditor samples n flagged threshold conflicts at random
# from a real library and labels each, blind to the book's claim, as
# a true contradiction (no value satisfies both rules) or not. The
# common belief is that at least half of flagged conflicts are true
# contradictions (H0: share >= 0.5). The book claims the share is
# small. Exact one-sided binomial test at alpha 0.05: reject H0 when
# the count is at most c, with c the largest value whose probability
# under 0.5 is at most 0.05. The table gives the smallest n with exact
# power of at least 0.80 and 0.90 for true shares 0.10 to 0.35.

import math


def cdf(c, n, p):
    if c < 0:
        return 0.0
    return math.fsum(math.comb(n, k) * p ** k * (1 - p) ** (n - k)
                     for k in range(c + 1))


def crit(n, alpha=0.05):
    c = -1
    while cdf(c + 1, n, 0.5) <= alpha:
        c += 1
    return c


def smallest_n(p1, target):
    for n in range(5, 400):
        c = crit(n)
        if c >= 0 and cdf(c, n, p1) >= target:
            return n, c
    return None, None


print("H0 share >= 0.5, one-sided exact binomial, alpha 0.05")
print("true share   n(80%)  reject if <=   n(90%)  reject if <=")
for p1 in (0.10, 0.20, 0.30, 0.35):
    n8, c8 = smallest_n(p1, 0.80)
    n9, c9 = smallest_n(p1, 0.90)
    print("   %.2f      %5d  %12d   %6d  %12d" % (p1, n8, c8, n9, c9))
n = 60
c = crit(n)
print("plan n = %d: reject H0 if at most %d are true contradictions"
      % (n, c))
print("   size under 0.5 = %.4f; power at 0.20 = %.4f, at 0.35 = %.4f"
      % (cdf(c, n, 0.5), cdf(c, n, 0.2), cdf(c, n, 0.35)))

唯一数据源与引用

范雅福(Fan, Y.). (2026). 《跨卷规则消解与公理归约、千条机器规则阈值冲突检测与最小原语集母典》第 8 章「实证闭环:剪枝器接入、月度归约与预注册检验」(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-258 · v1.0.0 · CC-BY-NC-ND-4.0

BibTeX
@book{fan2026258,
  author    = {Fan, Yafu},
  title     = {Cross-Volume Rule Resolution and Axiom Reduction: Detecting Threshold Conflicts among a Thousand Machine Rules and Finding a Minimal Primitive Set},
  year      = {2026},
  publisher = {5D Library},
  edition   = {1.0.0},
  note      = {DOI pending}
}

本书其他章节结论