关于「奇格常数:颈口有多窄,谱给出上下两道夹板」,《谱间隙与瓶颈割集》给出的定量结论是什么?

结论:第二章是主轴:奇格常数被谱从两侧夹住;剂量与反应扫描中,团内加边使平均度数升高,代数连通度却从 0.4384 降到 0.2042。

数据口径:书中参数为示意;全部计算由书中程序生成;本书是非个人化经营建议,任何生产系统或真实供应链的改动须由主事者人工复核放行。 第八章给出画图、体检、补边、复查的季度闭环,示意工作流三轮补边后代数连通度从 0.1824 升到 1.0000;蒸馏小工具「瓶颈割集体检器」;以及每组 74 个团队季度的预注册检验方案与推翻条件。

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: 0.2042, 0.4384

Scan A: one bridge, clique size k grows (degree up)
k n edges mean deg lambda2 h exact cut share
36 7 2.333 0.4384 0.3333 0.1429
48 13 3.250 0.3542 0.2500 0.0769
510 21 4.200 0.2984 0.2000 0.0476
612 31 5.167 0.2583 0.1667 0.0323
714 43 6.143 0.2280 0.1429 0.0233
816 57 7.125 0.2042 0.1250 0.0175
Scan B: k8 fixed, bridges b grow (edges at the bottleneck)
b edges mean deg lambda2 h exact lambda2 gain per edge
157 7.125 0.2042 0.1250 0.0000
258 7.250 0.4174 0.2500 0.2133
359 7.375 0.6411 0.3750 0.2185
460 7.500 0.8769 0.5000 0.2242
561 7.625 1.1270 0.6250 0.2307
662 7.750 1.3944 0.7500 0.2381
763 7.875 1.6834 0.8750 0.2465
864 8.000 2.0000 1.0000 0.2565
查看完整验算程序
# Chapter 2. p22: dose-response scan of the main claim on dumbbells:
# two cliques K_k joined by b disjoint bridges. Raising k raises every
# degree; raising b adds edges only where the bottleneck is.
# Model: lambda2 by Jacobi rotation; exact Cheeger constant h by subset
# enumeration (n <= 16); cut share = edges whose removal splits the
# network into halves, divided by all edges. Graph sizes are illustrative.
import math


def laplacian(n, edges):
    lap = [[0.0] * n for _ in range(n)]
    for i, j in edges:
        lap[i][j] -= 1.0
        lap[j][i] -= 1.0
        lap[i][i] += 1.0
        lap[j][j] += 1.0
    return lap


def jacobi(mat, tol=1e-22):
    n = len(mat)
    a = [row[:] for row in mat]
    v = [[1.0 if i == j else 0.0 for j in range(n)] for i in range(n)]
    sweeps = 0
    while True:
        off = math.fsum(a[i][j] ** 2 for i in range(n)
                        for j in range(i + 1, n))
        if off < tol or sweeps > 60:
            break
        sweeps += 1
        for p in range(n - 1):
            for q in range(p + 1, n):
                if abs(a[p][q]) < 1e-300:
                    continue
                th = (a[q][q] - a[p][p]) / (2.0 * a[p][q])
                t = math.copysign(1.0, th) / (abs(th) + math.hypot(th, 1.0))
                c = 1.0 / math.sqrt(t * t + 1.0)
                s = t * c
                for k in range(n):
                    akp, akq = a[k][p], a[k][q]
                    a[k][p] = c * akp - s * akq
                    a[k][q] = s * akp + c * akq
                for k in range(n):
                    apk, aqk = a[p][k], a[q][k]
                    a[p][k] = c * apk - s * aqk
                    a[q][k] = s * apk + c * aqk
                for k in range(n):
                    vkp, vkq = v[k][p], v[k][q]
                    v[k][p] = c * vkp - s * vkq
                    v[k][q] = s * vkp + c * vkq
    order = sorted(range(n), key=lambda i: a[i][i])
    vals = [a[i][i] for i in order]
    vecs = [[v[k][i] for k in range(n)] for i in order]
    return vals, vecs, sweeps


def cheeger_exact(n, edges):
    nbr = [0] * n
    for i, j in edges:
        nbr[i] |= 1 << j
        nbr[j] |= 1 << i
    deg = [bin(x).count("1") for x in nbr]
    bnd = [0] * (1 << n)
    best = None
    for mask in range(1, 1 << n):
        low = mask & -mask
        v = low.bit_length() - 1
        rest = mask ^ low
        bnd[mask] = bnd[rest] + deg[v] - 2 * bin(nbr[v] & rest).count("1")
        k = bin(mask).count("1")
        if 2 * k <= n:
            r = bnd[mask] / k
            if best is None or r < best:
                best = r
    return best


def dumbbell(k, b):
    left = [(a, c) for a in range(k) for c in range(a + 1, k)]
    right = [(a + k, c + k) for a, c in left]
    return left + right + [(k - 1 - i, k + i) for i in range(b)]


print("Scan A: one bridge, clique size k grows (degree up)")
print(" k   n   edges  mean deg  lambda2   h exact  cut share")
for k in range(3, 9):
    es = dumbbell(k, 1)
    n = 2 * k
    lam = jacobi(laplacian(n, es))[0][1]
    print("%2d  %2d   %4d    %6.3f   %7.4f   %7.4f   %7.4f"
          % (k, n, len(es), 2 * len(es) / n, lam, cheeger_exact(n, es),
             1 / len(es)))
print()
print("Scan B: k = 8 fixed, bridges b grow (edges at the bottleneck)")
print(" b   edges  mean deg  lambda2   h exact  lambda2 gain per edge")
base = None
for b in range(1, 9):
    es = dumbbell(8, b)
    lam = jacobi(laplacian(16, es))[0][1]
    gain = 0.0 if base is None else (lam - base) / (b - 1)
    if base is None:
        base = lam
    print("%2d   %4d    %6.3f   %7.4f   %7.4f   %7.4f"
          % (b, len(es), 2 * len(es) / 16, lam, cheeger_exact(16, es),
             gain))

唯一数据源与引用

范雅福(Fan, Y.). (2026). 《图拉普拉斯谱间隙、奇格常数与多智能体工作流及跨国供应链单点瓶颈割集母典》第 2 章「奇格常数:颈口有多窄,谱给出上下两道夹板」(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-243 · v1.0.0 · CC-BY-NC-ND-4.0

BibTeX
@book{fan2026243,
  author    = {Fan, Yafu},
  title     = {Graph-Laplacian Spectral Gaps and Cheeger Constants: Single-Point Bottleneck Cuts in Multi-Agent Workflows and Cross-Border Supply Chains},
  year      = {2026},
  publisher = {5D Library},
  edition   = {1.0.0},
  note      = {DOI pending}
}

本书其他章节结论