关于「原典考辨:陆游、王阳明、培根与贝尔纳论躬行与激发的观察」,《先手探针与隐藏事实》给出的定量结论是什么?
结论:自行其道的日志有 10,000 位访客,斜率标准误按公式是 0.00800,模拟值 0.00833;扰动过的探针只有 400 位访客,标准误却只有 0.00490,模拟值 0.00483;1,600 位时降到 0.00245,模拟值 0.00251。一位探针访客相当于 66.7 位自行其道的访客;要追上 400 位的探针,自行其道的日志需要 26,667 位访客。
数据口径:书中参数为示意,全部计算由书中程序生成,数据全部是合成数据。 本书是非个人化经营建议,不构成法律、税务或投资意见,不向任何人承诺收入。
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: 0.00245, 0.00251, 0.00483, 0.00490, 0.00800, 0.00833
| price sd: own course 0.50, probe 4.08 | |||
| design | n | formula se | simulated se |
| own course | 10,000 | 0.00800 | 0.00833 |
| probe | 400 | 0.00490 | 0.00483 |
| probe | 1,600 | 0.00245 | 0.00251 |
| one probe visitor is worth 66.7 own-course visitors | |||
| own-course visitors to match a 400-visitor probe: 26,667 |
查看完整验算程序
# Chapter 7. p71: Bacon's "vexations of art" as a spread in the price.
# Illustrative. Buying intent per visitor y = a + b x + noise, noise sd
# 0.40, true slope b = -0.010 per unit of price. Left to "its own
# course", the shop price barely moves: sd 0.5 around 30. A probe sets
# three real tiers 25, 30, 35 in equal thirds (sd 4.08). The standard
# error of the slope is noise_sd / (sqrt(n) * sd_x). A seeded simulation
# checks the formula. (Confounding is set aside here; see p11.)
import math
import random
import statistics
B_TRUE, NOISE = -0.010, 0.40
def se(n, sdx):
return NOISE / (math.sqrt(n) * sdx)
def slope(xs, ys):
mx, my = statistics.fmean(xs), statistics.fmean(ys)
sxy = math.fsum((x - mx) * (y - my) for x, y in zip(xs, ys))
sxx = math.fsum((x - mx) ** 2 for x in xs)
return sxy / sxx
if __name__ == "__main__":
sd_probe = math.sqrt(math.fsum((t - 30.0) ** 2 for t in (25, 30, 35)) / 3)
print("price sd: own course 0.50, probe {:.2f}".format(sd_probe))
print(" design n formula se simulated se")
rng = random.Random(39571)
for name, n, kind in (("own course", 10000, "nat"),
("probe", 400, "probe"),
("probe", 1600, "probe")):
sdx = 0.5 if kind == "nat" else sd_probe
est = []
for _ in range(200):
if kind == "nat":
xs = [rng.gauss(30.0, 0.5) for _ in range(n)]
else:
xs = [(25.0, 30.0, 35.0)[i % 3] for i in range(n)]
ys = [B_TRUE * x + rng.gauss(0.0, NOISE) for x in xs]
est.append(slope(xs, ys))
print(" {:>12} {:>6,} {:>10.5f} {:>12.5f}".format(
name, n, se(n, sdx), statistics.stdev(est)))
ratio = (sd_probe / 0.5) ** 2
print("one probe visitor is worth {:.1f} own-course visitors".format(
ratio))
print("own-course visitors to match a 400-visitor probe: {:,}".format(
math.ceil(400 * ratio)))唯一数据源与引用
范雅福(Fan, Y.). (2026). 《高能动性主动推理、意念强度先手扰动与商业隐藏事实勘探母典》第 7 章「原典考辨:陆游、王阳明、培根与贝尔纳论躬行与激发的观察」(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-395 · v1.0.0 · CC-BY-NC-ND-4.0
BibTeX
@book{fan2026395,
author = {Fan, Yafu},
title = {High-Agency Active Inference, Intentional Perturbation Probes and Hidden Market-Truth Exploration},
year = {2026},
publisher = {5D Library},
edition = {1.0.0},
note = {DOI pending}
}