关于「多人微力的有偏随机游走:漂移、噪声与主笔」,《扶乩意动与乩文溯源》给出的定量结论是什么?
结论:第三章把多手扶笔写成有偏随机游走,四只手在限时内到达目标的概率是 0.853,单手只有 0.120。
数据口径:立项蓝图把三层都写成既定事实;核对原始文献后,本书把后两层修正为有模型、待检验的假说,并说明法拉第 1853 年检验的是转桌而不是乩笔。 书中参数为示意,乩文语料全部为自拟合成示范;全部计算由书中程序生成。
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: 0.120, 0.853
| drift 0.02 and noise SD 0.30 per hand per step; target 10 units, 200 steps | |
| hands drift/step noise/step SNR/step P(reach) MC | |
| 1 | 0.020 0.300 0.0667 0.120 0.110 |
| 2 | 0.040 0.424 0.0943 0.484 0.460 |
| 4 | 0.080 0.600 0.1333 0.853 0.840 |
| 8 | 0.160 0.849 0.1886 0.986 0.982 |
| 16 | 0.320 1.200 0.2667 1.000 0.999 |
| lateral wander after T steps (SD, units): | |
| 1 hand(s) | 4.24 |
| 4 hand(s) | 8.49 |
| 16 hand(s) | 16.97 |
查看完整验算程序
# Chapter 3. p31: n micro-forces add up to a biased random walk
# Model: each of n hands pushes the pen with a shared drift mu toward
# the stroke the group expects plus independent 2-D Gaussian noise of
# SD sigma per axis per step. The pen therefore drifts n mu per step
# with noise sqrt(n) sigma: the signal-to-noise ratio per step grows as
# sqrt(n) mu / sigma. Along the target axis after T steps the position
# is Normal(n mu T, n sigma^2 T); P(reach D by T) uses the first-
# passage (inverse Gaussian) law. mu, sigma, D are illustrative units.
# 模型:多人微力叠加的有偏随机游走;参数为示意。
import math
import random
MU, SIGMA, D, T = 0.02, 0.30, 10.0, 200
def phi(x):
return 0.5 * (1.0 + math.erf(x / math.sqrt(2.0)))
def p_reach(v, s, d, t):
# first passage of Brownian motion with drift v, variance s^2
a = (v * t - d) / (s * math.sqrt(t))
b = (-v * t - d) / (s * math.sqrt(t))
return phi(a) + math.exp(2.0 * v * d / (s * s)) * phi(b)
def mc_reach(n, trials, rng):
hits = 0
for _ in range(trials):
x = 0.0
for _ in range(T):
x += n * MU + math.sqrt(n) * SIGMA * rng.gauss(0.0, 1.0)
if x >= D:
hits += 1
break
return hits / trials
if __name__ == "__main__":
rng = random.Random(20261112)
print("drift {:.2f} and noise SD {:.2f} per hand per step;"
" target {:.0f} units, {} steps".format(MU, SIGMA, D, T))
print(" hands drift/step noise/step SNR/step P(reach) MC")
for n in (1, 2, 4, 8, 16):
v, s = n * MU, math.sqrt(n) * SIGMA
print(" {:5d} {:10.3f} {:10.3f} {:8.4f} {:8.3f} {:.3f}"
.format(n, v, s, v / s, p_reach(v, s, D, T),
mc_reach(n, 2000, rng)))
print("lateral wander after T steps (SD, units):")
for n in (1, 4, 16):
print(" {:2d} hand(s): {:.2f}".format(
n, math.sqrt(n) * SIGMA * math.sqrt(T)))唯一数据源与引用
范雅福(Fan, Y.). (2026). 《扶乩意动效应微肌电、乩笔轨迹随机游走与乩文语言模型溯源母典》第 3 章「多人微力的有偏随机游走:漂移、噪声与主笔」(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-226 · v1.0.0 · CC-BY-NC-ND-4.0
BibTeX
@book{fan2026226,
author = {Fan, Yafu},
title = {Ideomotor Micro-Movements in Spirit Writing, Random-Walk Stylus Trajectories and Stylometric Attribution of Revealed Texts},
year = {2026},
publisher = {5D Library},
edition = {1.0.0},
note = {DOI pending}
}