关于「实践与证伪方案:农场主清单、文案检查器与预注册裂区试验」,《土壤电偶、铜杆天线与电耕试验》给出的定量结论是什么?
结论:第八章给出经营者清单、文案检查器与预注册裂区方案:检测一成效应约需 5320 平方米。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:书中参数为示意;全部计算由书中程序生成;本书不是农艺或投资建议,不评价任何品牌。 第一章用能斯特方程算出铜–锌电偶在示意土壤溶液中的开路电压约 1.0734 伏,并说明其能量来自锌的腐蚀。
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: 5320
| effect | blocks | t crit | power | plots | area m2 |
| 10% | 6 | 2.571 | 0.4015 | 36 | 2280 |
| 10% | 8 | 2.365 | 0.5445 | 48 | 3040 |
| 10% | 10 | 2.262 | 0.6680 | 60 | 3800 |
| 10% | 12 | 2.201 | 0.7570 | 72 | 4560 |
| 10% | 14 | 2.160 | 0.8375 | 84 | 5320 |
| 10% | 16 | 2.131 | 0.8840 | 96 | 6080 |
| 5% | 32 | 2.040 | 0.6275 | 192 | 12160 |
| 5% | 40 | 2.023 | 0.7300 | 240 | 15200 |
| 5% | 44 | 2.017 | 0.7795 | 264 | 16720 |
| 5% | 48 | 2.012 | 0.8195 | 288 | 18240 |
| 5% | 56 | 2.004 | 0.8880 | 336 | 21280 |
| SD of a block difference in main-plot means: 0.1178 | |||||
| effect 10%: normal formula 10.9 blocks; simulated power 0.8 at 14 | |||||
| effect 5%: normal formula 43.5 blocks; simulated power 0.8 at 48 |
查看完整验算程序
# Chapter 8. p81: sample size for the pre-registered split-plot trial.
# Main plots: rod or no rod; subplots: no zinc, zinc salt matched to
# the galvanic release of p52, double zinc. The rod effect is tested
# on main-plot means with r - 1 degrees of freedom (a paired t test
# on block differences, identical to F(1, r-1)). Effect to detect
# as below; main-plot SD 0.08 and subplot SD 0.04 (illustrative).
# Effects of 10 and 5 percent; power by fixed-seed Monte Carlo, 2000
# trials per row; 2.8016 = 1.9600 + 0.8416 for the normal formula.
# Field area uses 5 m x 6 m subplots and the 4 m guard strip of p63.
import math
import random
import statistics
def betacf(a, b, x):
qab, qap, qam = a + b, a + 1.0, a - 1.0
c, d = 1.0, 1.0 - qab * x / qap
d = 1.0 / (d if abs(d) > 1e-300 else 1e-300)
h = d
for m in range(1, 300):
m2 = 2 * m
for aa in (m * (b - m) * x / ((qam + m2) * (a + m2)),
-(a + m) * (qab + m) * x / ((a + m2) * (qap + m2))):
d = 1.0 + aa * d
d = 1.0 / (d if abs(d) > 1e-300 else 1e-300)
c = 1.0 + aa / c
c = c if abs(c) > 1e-300 else 1e-300
h *= d * c
if abs(d * c - 1.0) < 1e-14:
break
return h
def t_crit(df):
lo, hi = 0.0, 50.0
for _ in range(100):
t = (lo + hi) / 2.0
x = df / (df + t * t)
lb = (math.lgamma(df / 2.0 + 0.5) - math.lgamma(df / 2.0)
- math.lgamma(0.5) + df / 2.0 * math.log(x)
+ 0.5 * math.log(1.0 - x))
if x < (df / 2.0 + 1.0) / (df / 2.0 + 2.5):
up = 0.5 * math.exp(lb) * betacf(df / 2.0, 0.5, x) / (df / 2.0)
else:
up = 0.5 * (1.0 - math.exp(lb) * betacf(0.5, df / 2.0, 1.0 - x)
/ 0.5)
lo, hi = (t, hi) if up > 0.025 else (lo, t)
return (lo + hi) / 2.0
SW, SE, B = 0.08, 0.04, 3
rng = random.Random(20261112)
print("effect blocks t crit power plots area m2")
first = {}
for delta, grid in ((0.10, (6, 8, 10, 12, 14, 16)),
(0.05, (32, 40, 44, 48, 56))):
for r in grid:
crit = t_crit(r - 1)
hit = 0
for _ in range(2000):
d = []
for _ in range(r):
m1 = delta + rng.gauss(0.0, SW) + statistics.fmean(
rng.gauss(0.0, SE) for _ in range(B))
m0 = rng.gauss(0.0, SW) + statistics.fmean(
rng.gauss(0.0, SE) for _ in range(B))
d.append(m1 - m0)
t = statistics.fmean(d) / (statistics.stdev(d) / math.sqrt(r))
hit += abs(t) > crit
pw = hit / 2000
if pw >= 0.8 and delta not in first:
first[delta] = r
area = r * 2 * (B * 5.0 + 4.0) * (6.0 + 4.0)
print("{:5.0%} {:5d} {:6.3f} {:.4f} {:5d} {:7.0f}"
.format(delta, r, crit, pw, r * 2 * B, area))
sd_d = math.sqrt(2.0 * (SW ** 2 + SE ** 2 / B))
print("SD of a block difference in main-plot means: {:.4f}".format(sd_d))
for delta in (0.10, 0.05):
n = 2.8016 ** 2 * sd_d ** 2 / delta ** 2
print("effect {:.0%}: normal formula {:.1f} blocks; simulated power"
" 0.8 at {}".format(delta, n, first.get(delta)))唯一数据源与引用
范雅福(Fan, Y.). (2026). 《土壤电化学能斯特电势、铜棒无源天线纳瓦级收集极限与电耕增产声称的田间区组设计》第 8 章「实践与证伪方案:农场主清单、文案检查器与预注册裂区试验」(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-170 · v1.0.0 · CC-BY-NC-ND-4.0
BibTeX
@book{fan2026170,
author = {Fan, Yafu},
title = {Nernst Potentials in Soil Electrochemistry, Nanowatt Harvest Limits of Passive Copper Antennas and Field Block Designs for Electroculture Yield Claims},
year = {2026},
publisher = {5D Library},
edition = {1.0.0},
note = {DOI pending}
}