关于「实证闭环:周度调货单、影子价表与人工放行」,《最优传输与奢品调货》给出的定量结论是什么?
结论:第八章给出周度调货流程、跨店调货最优搬运器与预注册检验方案:成本检验需要 197 对周,十二个区域约 34 周。书中以标注框区分已发表的证据、可运行的 Python 计算(18 段程序)、原典与作者的假说。
数据口径:书中的门店、城市坐标、运价、库存、毛利与概率均为示意;全部计算由书中程序生成;本书是非个人化经营建议,书中小工具给出的每一条调拨都须由经营者人工放行。 第三章把位势读成地点价格:计入缺货毛利损失后,示意网络的最优总成本是 7322 元,「缺货店优先补」是 11003 元;每家店多一件货的影子价格都挤在边际缺货店的 1320 元附近,缺一件损失最大的店只值 1253 元,而影子价格是局部斜率,同一家店多 8 件只值 5709 元。
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: 197, 34
| z(0.975) | 1.9600, z(0.80) = 0.8416 |
| test 1: paired weeks needed (baseline 10,000 yuan/week) | |
| sd_d \ delta | 300 500 800 1200 |
| 1500 | 197 71 28 13 |
| 2500 | 546 197 77 35 |
| 4000 | 1396 503 197 88 |
| test 2: received units needed to detect correlation r | |
| r | 0.15 n = 347 |
| r | 0.20 n = 194 |
| r | 0.30 n = 85 |
| r | 0.40 n = 47 |
| plan: sd_d 2500, delta 500 | 197 pairs; 12 regions -> 17 cycles |
| of two weeks each | 34 weeks, 204 pairs |
查看完整验算程序
# Chapter 8. p82: sample sizes for the pre-registered tests.
# Test 1 (cost): alternate weeks, region by region, between the
# shortage-first rule and the optimiser; outcome = weekly freight plus
# realised lost margin; paired difference per region-pair of weeks.
# n = ((z_{1-a/2} + z_{1-b}) * sd_d / delta)^2, two-sided a = 0.05,
# power 0.80; sd_d and delta are illustrative, in yuan per week.
# Test 2 (prices): does the shadow value predict the realised gross
# margin of a received unit? Fisher z: n = ((z_a + z_b) / atanh r)^2 + 3.
import math
from statistics import NormalDist
Z = NormalDist()
za = Z.inv_cdf(1 - 0.05 / 2)
zb = Z.inv_cdf(0.80)
print("z(0.975) = %.4f, z(0.80) = %.4f" % (za, zb))
print("test 1: paired weeks needed (baseline 10,000 yuan/week)")
print("sd_d \ delta 300 500 800 1200")
for sd in (1500, 2500, 4000):
row = [math.ceil(((za + zb) * sd / dl) ** 2)
for dl in (300, 500, 800, 1200)]
print("%6d " % sd + "".join("%6d" % n for n in row))
print("test 2: received units needed to detect correlation r")
for r in (0.15, 0.2, 0.3, 0.4):
n = math.ceil(((za + zb) / math.atanh(r)) ** 2 + 3)
print(" r = %.2f n = %d" % (r, n))
need = math.ceil(((za + zb) * 2500 / 500) ** 2)
cyc = math.ceil(need / 12)
print("plan: sd_d 2500, delta 500 -> %d pairs; 12 regions -> %d cycles"
% (need, cyc))
print(" of two weeks each = %d weeks, %d pairs" % (2 * cyc, 12 * cyc))唯一数据源与引用
范雅福(Fan, Y.). (2026). 《最优传输康托罗维奇对偶、跨区域奢品调货最小搬运代价与客群迁移矩阵母典》第 8 章「实证闭环:周度调货单、影子价表与人工放行」(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-233 · v1.0.0 · CC-BY-NC-ND-4.0
BibTeX
@book{fan2026233,
author = {Fan, Yafu},
title = {Optimal Transport and Kantorovich Duality: Minimum-Cost Rebalancing of Luxury Stock across Regions and Client-Migration Matrices},
year = {2026},
publisher = {5D Library},
edition = {1.0.0},
note = {DOI pending}
}