关于「实践:样宅 A、样盘 A 与一份预先登记的检验」,《玄空声光场与汉堡择时》给出的定量结论是什么?
结论:第八章用样宅 A 与样盘 A 走完全程,并给出预先登记的检验:两个主要睡眠结局、效应量 0.3 时需要 109 户;只申报三条轴线时,0.25 度的核心窗口只覆盖 0.0328 的周,而把 156 种接触全部算上则覆盖 0.5796。书中以标注框区分已发表的证据、可运行的 Python 计算、原典与作者的假说。
数据口径:两套传统按原有算法写准、算尽;三者的组合是本书的处方,也是一个可检验的假说,不是已被证实的效果。 全部计算由书中程序生成;本书不构成医疗、建筑安全、投资或心理方面的专业建议。
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: 0.0328, 0.25, 0.3, 0.5796, 109, 156
| test 1, homes needed (crossover) | |
| d_z alpha 0.05 alpha 0.025 | |
| 0.5 | 34 41 |
| 0.3 | 90 109 |
| 0.2 | 199 241 |
| test 2, people needed (26 logged weeks each) | |
| d icc 0.3 icc 0.5 icc 0.7 | |
| 0.3 | 30 47 65 |
| 0.2 | 66 104 142 |
| share of weeks inside at least one window | |
| all 156 contacts orb 1.50 | 0.9945 orb 1.00: 0.9688 orb 0.25: 0.5796 |
| 3 declared axes orb 1.50 | 0.1813 orb 1.00: 0.1248 orb 0.25: 0.0328 |
查看完整验算程序
# Chapter 8. p07: Sample sizes for the pre-registered test of the book.
# Test 1 (space). Crossover, each home lives 4 weeks under the chart-dealt
# zone uses and 4 weeks under a sham dealing (the chart of the opposite
# sitting mountain), both with the SAME light and sound targets, order
# randomised. Outcomes by wrist actigraphy (sleep efficiency, sleep onset
# latency) and a pre-declared task-output count. Paired effect size d_z;
# n = ((z_a + z_b) / d)^2 + z_a^2 / 2, two-sided alpha 0.05 (or 0.025
# for two primary outcomes), power 0.80.
# Test 2 (time). Each person's weekly output inside the core windows of
# their own chart against the windows of a chart drawn from another
# participant (blind, as in Carlson 1985). Weeks per person w, intraclass
# correlation icc; design effect 1 + (w - 1) icc on the person count.
# Last part: with k directed contacts each open 2 orb / 90 of the time,
# the share of weeks under at least one window is 1 - exp(-2 k orb / 90).
from math import ceil, exp
Z = {0.05: 1.9600, 0.025: 2.2414}
ZB = 0.8416
def n_paired(d, alpha):
za = Z[alpha]
return ceil(((za + ZB) / d) ** 2 + za ** 2 / 2)
def n_people(d, weeks, icc, alpha=0.05):
deff = 1 + (weeks - 1) * icc
return ceil(n_paired(d, alpha) * deff / weeks)
if __name__ == "__main__":
print(" test 1, homes needed (crossover)")
print(" d_z alpha 0.05 alpha 0.025")
for d in (0.5, 0.3, 0.2):
print(" {:.1f} {:10d} {:11d}".format(d, n_paired(d, 0.05), n_paired(d, 0.025)))
print("\n test 2, people needed (26 logged weeks each)")
print(" d icc 0.3 icc 0.5 icc 0.7")
for d in (0.3, 0.2):
print(" {:.1f} {:7d} {:7d} {:7d}".format(
d, *[n_people(d, 26, icc) for icc in (0.3, 0.5, 0.7)]))
print("\n share of weeks inside at least one window")
for k, label in ((156, "all 156 contacts"), (6, "3 declared axes")):
print(" {:16s} ".format(label) + " ".join(
"orb {:.2f}: {:.4f}".format(o, 1 - exp(-2 * k * o / 90))
for o in (1.5, 1.0, 0.25)))唯一数据源与引用
范雅福(Fan, Y.). (2026). 《玄空飞星空间声光场与汉堡中点择时母典》第 8 章「实践:样宅 A、样盘 A 与一份预先登记的检验」(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-122 · v1.0.0 · CC-BY-NC-ND-4.0
BibTeX
@book{fan2026122,
author = {Fan, Yafu},
title = {The Xuankong Flying-Star Sound-and-Light Field and Hamburg Midpoint Timing Codex},
year = {2026},
publisher = {5D Library},
edition = {1.0.0},
note = {DOI pending}
}