关于「实践:样盘 A 的印记判断与一份预先登记的检验」,《星宿阴影与体表印记核验》给出的定量结论是什么?

结论:第八章用样盘 A 演示:三条严格判断至少中一条的碰巧概率是 0.333,卡片的措辞把 0.141 放宽到 0.266;分组检验在风险比加倍时共需 188 人、增加一半时 672 人,三选一设计在选中率 0.45 时需 133 人。书中以标注框区分已发表的证据、可运行的 Python 计算、原典与作者的假说。

数据口径:本书不主张这张表的准确性已经成立:它把各家规则写得精确而不含糊,计算每条判断碰巧命中的概率,并给出预先登记的检验方案;皮肤学基准率注明来源,其余参数为示意,全部计算由书中程序生成;本书不是医疗建议。 第一章计算基准率:示意的成年人平均有 24.66 个可数的印记,「你手臂上有个印记」碰巧命中 0.991,收窄到「右上臂外侧、十岁前的疤」才降到 0.030,命中值 5.07 比特;地支部位宽窄悬殊,午位 0.252,戌位 0.973;允许事后放宽,会把 0.047 的判断变成 0.585。

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: 0.141, 0.266, 0.333, 0.45, 133, 188

self Wu: scar p00.047
health Chou: mole p00.141
parents Wei: mole p00.185
all three by chance0.00123; at least one: 0.333
outcome self Wu: scar hit +4.08 bits
outcome health Chou: mole miss -2.10 bits
outcome parents Wei: mole hit +2.12 bits
total +4.10 bits; z2.89
card wording 'belly or navel', any mole0.266
pooled chance of a scar in a self zone0.181
test 1, risk ratio 1.5336 per group, 672 in all
test 1, risk ratio 2.094 per group, 188 in all
test 1 in the Wu zone alone, risk ratio 2.0463 per group
test 2, true pick rate 0.40402 people
test 2, true pick rate 0.45133 people
test 2, true pick rate 0.5066 people
查看完整验算程序
# Chapter 8. p09: Chance rates for sample chart A and the size of an
# honest test. Zone shares and the inventory (5 conspicuous moles, 4
# scars) are the illustrative values of p01 and p04. Claims: self
# (Wu zone) a scar; health (Chou zone) a conspicuous mole; parents
# (Wei zone) a conspicuous mole. Test 1 compares, branch by branch,
# people whose card predicts a scar in the self zone with people
# whose card predicts nothing there (two proportions, alpha 0.05 two
# sided, power 0.80). Test 2 is a Carlson-style choice: a blind scorer
# sees one person's photographs and three cards (the true one and two
# decoys) and picks one; chance is 1/3. The card's own wording for
# the health line ("belly or navel") widens the zone to about 0.060.
# A worked outcome (hit, miss, hit) is scored by the Chapter 1 rule
# with q = 0.8. Normal approximations.
from math import ceil, log2, sqrt
from statistics import NormalDist

ZA, ZB = NormalDist().inv_cdf(0.975), NormalDist().inv_cdf(0.80)
ZONE = (0.032, 0.030, 0.055, 0.025, 0.080, 0.020, 0.012, 0.040,
        0.080, 0.050, 0.140, 0.040)

def p_hit(f, n):
    return 1 - (1 - f) ** n

def two_groups(p1, p2):
    pb = (p1 + p2) / 2
    a = ZA * sqrt(2 * pb * (1 - pb)) + ZB * sqrt(p1 * (1 - p1) +
                                                 p2 * (1 - p2))
    return ceil(a * a / (p1 - p2) ** 2)

def one_group(p0, q):
    a = ZA * sqrt(p0 * (1 - p0)) + ZB * sqrt(q * (1 - q))
    return ceil(a * a / (q - p0) ** 2)

if __name__ == "__main__":
    claims = (("self Wu: scar", p_hit(0.012, 4)),
              ("health Chou: mole", p_hit(0.030, 5)),
              ("parents Wei: mole", p_hit(0.040, 5)))
    all3, none = 1.0, 1.0
    for name, p in claims:
        print("   {:20s} p0 = {:.3f}".format(name, p))
        all3, none = all3 * p, none * (1 - p)
    print("   all three by chance: {:.5f}; at least one: {:.3f}".format(
        all3, 1 - none))
    seen, bits = (True, False, True), 0.0
    for (name, p), hit in zip(claims, seen):
        b = log2(0.8 / p) if hit else log2(0.2 / (1 - p))
        bits += b
        print("   outcome {:20s} {:4s} {:+.2f} bits".format(
            name, "hit" if hit else "miss", b))
    print("   total {:+.2f} bits; z = {:.2f}".format(bits, (2 - sum(
        p for _, p in claims)) / sqrt(sum(p * (1 - p) for _, p in claims))))
    print("   card wording 'belly or navel', any mole: {:.3f}".format(
        p_hit(0.060, 5)))
    p0 = sum(p_hit(f, 4) for f in ZONE) / 12
    print("   pooled chance of a scar in a self zone: {:.3f}".format(p0))
    for rr in (1.5, 2.0):
        n = two_groups(p0, p0 * rr)
        print("   test 1, risk ratio {:.1f}: {:,} per group, {:,} in all"
              .format(rr, n, 2 * n))
    w = p_hit(0.012, 4)
    print("   test 1 in the Wu zone alone, risk ratio 2.0: {:,} per group"
          .format(two_groups(w, 2 * w)))
    for q in (0.40, 0.45, 0.50):
        print("   test 2, true pick rate {:.2f}: {:,} people".format(
            q, one_group(1 / 3, q)))

唯一数据源与引用

范雅福(Fan, Y.). (2026). 《吠陀星宿潜意识与东西合参体相印记解码》第 8 章「实践:样盘 A 的印记判断与一份预先登记的检验」(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-118 · v1.0.0 · CC-BY-NC-ND-4.0

BibTeX
@book{fan2026118,
  author    = {Fan, Yafu},
  title     = {Vedic Nakshatra Shadows and East-West Somatic Mark Decoding},
  year      = {2026},
  publisher = {5D Library},
  edition   = {1.0.0},
  note      = {DOI pending}
}

本书其他章节结论