5D Library › 古典卜卦尊贵度积分、恒星二十七宿岁差与维姆肖塔里运限链 · English
第二章 互溶、定位星映射与截夺的首达竞争
定位星映射
每颗行星落在某个星座,那个星座的庙主就是它的「定位星」。于是七颗行星构成一个映射 :行星 其所在星座的庙主。这是一个七元集合到自身的映射,它的函数图由若干个环和挂在环上的树组成。
- 行星在自己的庙宫,就是 的不动点(长度为 1 的环);
- 两颗行星互在对方的庙宫,即「互溶」(mutual reception),就是 的 2-循环,也就是交换两者的对换;
- 若 唯一的环是一个不动点,那颗行星就是全盘的「最终定位星」。
推导
# Chapter 2. p03: Mutual reception as 2-cycles of the dispositor map.
# Each planet is "disposed" by the domicile ruler of the sign it is in:
# f(planet) = ruler(sign(planet)), a map of the 7 planets to themselves.
# A mutual reception (by domicile) is a 2-cycle of f; a planet in its own
# domicile is a fixed point; a "final dispositor" exists when the only
# cycle of f is one fixed point. Exact pair rates for uniform independent
# signs, then the real sky (p02), where Mercury and Venus hug the Sun.
import importlib.util
import random
from fractions import Fraction
def load(name):
spec = importlib.util.spec_from_file_location(name, name + ".py")
m = importlib.util.module_from_spec(spec)
spec.loader.exec_module(m)
return m
dig = load("p01_dignity_tables")
eph = load("p02_ephemeris")
RULER = {s: p for p, ss in dig.DOM.items() for s in ss}
def cycles(signs):
f = {p: RULER[signs[p]] for p in dig.PL}
mutual = sum(1 for i, a in enumerate(dig.PL) for b in dig.PL[i + 1:]
if f[a] == b and f[b] == a)
fixed = [p for p in dig.PL if f[p] == p]
seen, cyc = set(), []
for p in dig.PL:
path, x = [], p
while x not in seen and x not in path:
path.append(x)
x = f[x]
if x in path:
cyc.append(path[path.index(x):])
seen |= set(path)
final = len(cyc) == 1 and len(cyc[0]) == 1
return mutual, len(fixed), final
def rates(samples):
n = len(samples)
return (sum(m > 0 for m, _, _ in samples) / n,
sum(m for m, _, _ in samples) / n,
sum(f > 0 for _, f, _ in samples) / n,
sum(fd for _, _, fd in samples) / n)
if __name__ == "__main__":
n = {p: len(dig.DOM[p]) for p in dig.PL}
exp = sum(Fraction(n[a] * n[b], 144) for i, a in enumerate(dig.PL)
for b in dig.PL[i + 1:])
print(" uniform signs: expected mutual receptions {} = {:.4f}".format(
exp, float(exp)))
rng = random.Random(73003)
uni = [cycles({p: rng.randrange(12) for p in dig.PL})
for _ in range(100000)]
sky = []
for _ in range(40000):
q = eph.positions(rng.uniform(2415020.5, 2469807.5))
sky.append(cycles({p: int(q[p] // 30) for p in dig.PL}))
print(" model P(>=1 mutual) E[mutual] P(own domicile)"
" P(final disp.)")
for name, s in (("uniform", uni), ("sky", sky)):
print(" {:8s} {:13.4f} {:9.4f} {:15.4f} {:14.4f}".format(
name, *rates(s)))
# uniform signs: expected mutual receptions 61/144 = 0.4236
# model P(>=1 mutual) E[mutual] P(own domicile) P(final disp.)
# uniform 0.3815 0.4218 0.6610 0.1542
# sky 0.3596 0.3859 0.6669 0.1942互溶的基率
若七颗行星的星座独立均匀,一对行星 、 互溶的概率是 ,其中 为各自的庙宫数(太阳、月亮各 1,其余各 2)。于是互溶对数的期望为 61/144,约 0.4236。模拟给出:至少有一对互溶的概率为 0.3815,至少一颗行星入庙的概率为 0.6610,存在最终定位星的概率为 0.1542。
用 1900 到 2049 年的真实位置计算,至少一对互溶的概率为 0.3596,入庙 0.6669,最终定位星 0.1942。真实天空中水星、金星与太阳靠得很近,它们之间(例如水星在狮子座、太阳在双子座或处女座)的互溶更难出现,所以互溶略少;最终定位星则略多。
注
「互溶」并不罕见。 大约每三张盘就有一张至少有一对庙宫互溶。若再把旺、三分、界、面的互容也算进来,比例会更高。互溶本身不是一张盘的特别之处。
入相位与截夺:首达竞争
两颗行星的黄经为 、,日行速度为 、。若它们按直线运动,形成角度 (0、60、90、120、180 度)的精确相位所需的时间是方程
的最小正根;只要这个根为正,相位就是「入相位」(applying)。卜卦占星里,若月亮正入相位于某颗主星 ,而在它完成之前,第三颗行星先与 成相位,就叫「截夺」(prohibition)。这是一个首达时间的竞争:谁先到达自己的目标角度。
推导
# Chapter 2. p04: Applying aspects as a first-passage race.
# Two bodies with longitudes l1, l2 and daily motions w1, w2 form an
# aspect of angle a (0, 60, 90, 120, 180) when l1 - l2 = +/- a (mod 360).
# Moving linearly, the time to the next exact aspect is the smallest
# positive t with (l1 - l2) + (w1 - w2) t = +/- a (mod 360); the aspect is
# "applying" while that t is positive and the separation is closing.
# Question: the Moon applies to a significator X; before it perfects,
# does a third planet Y perfect an aspect to X first (prohibition)?
import importlib.util
import random
def load(name):
spec = importlib.util.spec_from_file_location(name, name + ".py")
m = importlib.util.module_from_spec(spec)
spec.loader.exec_module(m)
return m
eph = load("p02_ephemeris")
ASPECTS = (0, 60, 90, 120, 180)
BODIES = ("Sun", "Mercury", "Venus", "Mars", "Jupiter", "Saturn")
def next_perfection(l1, w1, l2, w2, horizon):
dw = w1 - w2
if abs(dw) < 1e-9:
return None
d = l1 - l2
best = None
for a in ASPECTS:
for target in {a, -a}:
t = ((target - d) % 360) / dw if dw > 0 else \
-((d - target) % 360) / dw
if 0 < t <= horizon and (best is None or t < best):
best = t
return best
if __name__ == "__main__":
rng = random.Random(73004)
N = 20000
stats = {x: [0, 0, 0.0] for x in BODIES}
for _ in range(N):
jd = rng.uniform(2415020.5, 2469807.5)
pos = eph.positions(jd)
w = {b: eph.speed(jd, b) for b in BODIES + ("Moon",)}
for x in BODIES:
tm = next_perfection(pos["Moon"], w["Moon"], pos[x], w[x], 30)
first = min((t for y in BODIES if y != x for t in [
next_perfection(pos[y], w[y], pos[x], w[x], tm)]
if t is not None), default=None)
stats[x][0] += 1
stats[x][1] += first is not None
stats[x][2] += tm
print(" significator Moon perfects in (days) prohibited first")
for x in BODIES:
n, k, t = stats[x]
print(" {:12s} {:23.2f} {:16.4f}".format(x, t / n, k / n))
print(" (third bodies: the other five of Sun, Mercury ... Saturn)")
# significator Moon perfects in (days) prohibited first
# Sun 2.04 0.1330
# Mercury 2.06 0.1886
# Venus 2.06 0.1409
# Mars 1.98 0.1283
# Jupiter 1.93 0.1442
# Saturn 1.92 0.1515
# (third bodies: the other five of Sun, Mercury ... Saturn)在 1900 到 2049 年的随机时刻,月亮完成下一个对某颗行星的托勒密相位平均需要 1.92 到 2.06 天。在这段时间里,被其他五颗行星(太阳、水星至土星中除 以外的)之一抢先与 成相位的比例,从火星作主星时的 0.1283 到水星作主星时的 0.1886 不等。水星最容易被截夺,因为它自身运动快、又常与太阳、金星同处一区,与它们的相对运动使精确相位频繁出现。
假说 · 尚未验证
作者的假说(尚未验证): 截夺在卜卦判断中的出现频率主要由上述运动学基率决定,与问题的内容无关。什么会证明它错: 若在同一批时刻上,不同类别问题的星盘中截夺出现的比例有显著差异,这一假说就不成立(因为星盘只由时刻决定,这一差异只能来自选择问题的时刻)。
小结
互溶是定位星映射的 2-循环,在均匀模型下互溶对数的期望为 61/144,至少一对的概率为 0.3815,真实天空中为 0.3596。截夺是首达竞争:月亮入相位期间被第三颗行星抢先的比例约为一成三到一成九。
Fan, Y. (2026). Horary Dignity Scoring, Sidereal Nakshatra Precession and Vimshottari Dasha Chains: Reception Matrix Algebra, Ayanamsha Drift Kinetics and Non-Uniform Planetary Period Partitions [古典卜卦尊贵度积分、恒星二十七宿岁差与维姆肖塔里运限链] (Version 1.0.0). 5D Library. https://doi.org/10.5281/zenodo.24090195 · Licensed under Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International (CC BY-NC-ND 4.0).