关于「批量越大越赚钱吗:报童临界分位与零库存的盈亏临界」,《重磅棉力学、印花耐洗与小批量经济学》给出的定量结论是什么?
结论:第六章用报童模型说明批量决策取决于临界分位而非单价曲线:盈亏临界变异系数随均值需求变化,示意参数下均值 120 件约为 0.593、40 件约为 0.392,为凑价格档做满 250 件的方案单价最低、利润也最低。
数据口径:书中的布料、洗法、成本与需求参数均为示意,全部计算由书中程序生成;本书是非个人化经营建议,耐久结论一律写成条件化、可检验的预测,不作任何绝对化承诺,洗涤与尺寸标准只引编号与适用范围。 第六章用报童模型说明批量决策取决于临界分位而非单价曲线:盈亏临界变异系数随均值需求变化,示意参数下均值 120 件约为 0.593、40 件约为 0.392,为凑价格档做满 250 件的方案单价最低、利润也最低。
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: 0.392, 0.593, 120, 40
| mu | Poisson floor | CV* |
|---|---|---|
| 15 | 0.258 | any |
| 20 | 0.224 | any |
| 30 | 0.183 | 0.288 |
| 40 | 0.158 | 0.392 |
| 60 | 0.129 | 0.493 |
| 120 | 0.091 | 0.593 |
| 240 | 0.065 | 0.643 |
| check mu 40, CV 0.8: Q* = 61, exact 1417.6, simulated 1427.5 | ||
| on demand at mu 40: 1878.0 |
查看完整验算程序
# Chapter 6. p62: the break-even CV with discrete, skewed demand.
# Demand per design is negative binomial (Poisson rate drawn from a
# gamma), so CV^2 = 1 / mu + 1 / r and small designs are never below
# the Poisson floor 1 / sqrt(mu). Bulk profit is computed exactly from
# the pmf at the newsvendor order; on-demand as in p61. CV* is found by
# bisection; a fixed-seed Monte Carlo checks one point. Illustrative.
import math
import random
P, CB, V, KB = 89.0, 22.0, 5.0, 400.0
COD, KOD, ALPHA = 38.0, 60.0, 0.05
CF = (P - CB) / (P - V)
def pmf(mu, cv):
r = 1.0 / (cv * cv - 1.0 / mu)
q = r / (r + mu)
out, pk, k = [], q ** r, 0
while k < 2 * mu + 50 or pk > 1e-13:
out.append(pk)
pk *= (k + r) / (k + 1) * (1 - q)
k += 1
return out
def bulk(mu, cv):
f = pmf(mu, cv)
lo, hi = 0, len(f) - 1 # smallest q with F(q) >= CF
while lo < hi:
mid = (lo + hi) // 2
if math.fsum(f[:mid + 1]) >= CF:
hi = mid
else:
lo = mid + 1
q = lo
sold = math.fsum(min(k, q) * pk for k, pk in enumerate(f))
left = math.fsum(max(q - k, 0) * pk for k, pk in enumerate(f))
return q, P * sold + V * left - CB * q - KB
def od(mu):
return (P - COD) * (1 - ALPHA) * mu - KOD
def cv_star(mu):
lo = 1.0 / math.sqrt(mu) + 1e-6
if bulk(mu, lo)[1] < od(mu):
return "any"
hi = 3.0
if bulk(mu, hi)[1] > od(mu):
return "> 3"
for _ in range(50):
mid = (lo + hi) / 2
if bulk(mu, mid)[1] > od(mu):
lo = mid
else:
hi = mid
return "%.3f" % ((lo + hi) / 2)
print("mu Poisson floor CV*")
for mu in (15, 20, 30, 40, 60, 120, 240):
print("%3d %.3f %s" % (mu, 1 / math.sqrt(mu), cv_star(mu)))
rng = random.Random(20261120)
mu, cv = 40, 0.8
q, exact = bulk(mu, cv)
r = 1.0 / (cv * cv - 1.0 / mu)
prof = []
for _ in range(20000):
lam = rng.gammavariate(r, mu / r)
d, t = 0, rng.expovariate(1.0)
while t < lam:
d += 1
t += rng.expovariate(1.0)
prof.append(P * min(d, q) + V * max(q - d, 0) - CB * q - KB)
print("check mu %d, CV %.1f: Q* = %d, exact %.1f, simulated %.1f" % (
mu, cv, q, exact, math.fsum(prof) / len(prof)))
print("on demand at mu %d: %.1f" % (mu, od(mu)))唯一数据源与引用
范雅福(Fan, Y.). (2026). 《重磅棉织物克重力学、热转印涂层耐洗疲劳与小批量柔性定制经济学母典》第 6 章「批量越大越赚钱吗:报童临界分位与零库存的盈亏临界」(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-337 · v1.0.0 · CC-BY-NC-ND-4.0
BibTeX
@book{fan2026337,
author = {Fan, Yafu},
title = {Heavyweight Cotton Areal-Density Mechanics, Heat-Transfer Coating Wash-Cycle Durability and Micro-Batch Custom Apparel Economics},
year = {2026},
publisher = {5D Library},
edition = {1.0.0},
note = {DOI pending}
}