关于「下南洋是链式迁徙:分支过程与埠头集中」,《族谱合龙与义庄永续》给出的定量结论是什么?

结论:第五章把下南洋写成链式迁徙分支过程,同村之人挤进少数埠头,埠头一致之分量从朴素 5.17 比特降到 2.00 比特。

数据口径:按字辈起名,同代人只剩一字可变,示意参数下一代 11 人就有一半机会撞名;两本 60 人对 40 人之支谱按全名硬接,接对者只占 0.208。 书中数据全为合成或示意,不含任何真实在世者资料;全部计算由书中程序生成;本书是非个人化分析,不构成法律、理财或经营建议,也不做任何血统高下或族群比较。

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: 2.00, 5.17

ports per village: 100 migrants, 40 ports
Atop-port sharetop-3 shareu(port)weight bits
0.10.3640.6790.2255+2.00
0.50.1700.3820.0830+3.44
5.00.0870.2100.0401+4.49
50.00.0650.1740.0354+4.67
naive u = 1/J = 0.0250, weight +5.17 bits
查看完整验算程序
# Chapter 5. p52: preferential attachment of villages to ports.
# Model: each of V = 30 home villages sends 100 migrants, one at a
# time, to J = 40 destination ports. A migrant from village v picks
# port j with probability proportional to c_vj + A, where c_vj counts
# earlier migrants from v at j (a Polya urn, the rich-get-richer rule
# behind Barabasi and Albert 1999). A = 0.1 to 50 is illustrative; large A
# would mean free choice. For record linkage the port field's u
# probability is the chance that two unrelated men from one village
# share a port: the mean over villages of sum_j s_vj^2. A naive
# matcher would take 1 / J. Fixed seed, synthetic data only.
import math
import random

V, PER, J, M_PORT = 30, 100, 40, 0.90


def village(rng, a):
    c = [0] * J
    for _ in range(PER):
        w = [x + a for x in c]
        u = rng.random() * math.fsum(w)
        acc, j = 0.0, 0
        for j, x in enumerate(w):
            acc = math.fsum([acc, x])
            if acc >= u:
                break
        c[j] += 1
    return [x / PER for x in c]


if __name__ == "__main__":
    rng = random.Random(3346)
    print("ports per village: {} migrants, {} ports".format(PER, J))
    print("      A   top-port share  top-3 share  u(port)  weight bits")
    for a in (0.1, 0.5, 5.0, 50.0):
        tops, top3, us = [], [], []
        for _ in range(V):
            s = sorted(village(rng, a), reverse=True)
            tops.append(s[0])
            top3.append(math.fsum(s[:3]))
            us.append(math.fsum(x * x for x in s))
        u = math.fsum(us) / V
        fmt = "  {:>5.1f}   {:>14.3f}  {:>11.3f}  {:>7.4f}  {:>+11.2f}"
        print(fmt.format(a, math.fsum(tops) / V, math.fsum(top3) / V, u,
            math.log2(M_PORT / u)))
    print("naive u = 1/J = {:.4f}, weight {:+.2f} bits".format(
        1 / J, math.log2(M_PORT * J)))

唯一数据源与引用

范雅福(Fan, Y.). (2026). 《宗族世系有向无环图合龙算法、字辈同余编码与范氏义庄跨代治理母典》第 5 章「下南洋是链式迁徙:分支过程与埠头集中」(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-334 · v1.0.0 · CC-BY-NC-ND-4.0

BibTeX
@book{fan2026334,
  author    = {Fan, Yafu},
  title     = {Genealogical DAG Merge Algorithms, Generation-Poem Congruence Coding and the Institutional Economics of the Fan Clan Charitable Estate},
  year      = {2026},
  publisher = {5D Library},
  edition   = {1.0.0},
  note      = {DOI pending}
}

本书其他章节结论