Skip to content

Commit e0f7f21

Browse files
HUF discovery
1 parent 9f88d35 commit e0f7f21

2 files changed

Lines changed: 81 additions & 10 deletions

File tree

README.md

Lines changed: 26 additions & 10 deletions
Original file line numberDiff line numberDiff line change
@@ -1,26 +1,40 @@
11
# Higgins Unity Framework (HUF)
22

3-
*Compositional structure as an open-loop monitoring signal.*
3+
**An independent research project in Compositional Data Analysis (CoDa) and entropy-invariant monitoring on the simplex.** This is a scientific research repository — not a game engine, not a Unity plugin, not a software library.
44

5-
HUF proposes that **composition** — the internal proportional balance of a system's parts — can be monitored as a primary observable alongside magnitude, identity, and trend. The instrument reads. The human expert decides.
5+
*Author: Peter Higgins, Rogue Wave Audio, Markham, Ontario, Canada*
6+
7+
**Core discovery:** The Entropy-Invariant Time Transformer (EITT) — Shannon entropy of compositional time series is empirically near-invariant under geometric-mean decimation. Measured at **0.18% variation** across a 341:1 compression ratio. Validated across energy systems, chemistry (500,000 data points), hardware reliability, and climate scenarios.
8+
9+
**Disciplines:** Compositional Data Analysis, Shannon entropy, Aitchison geometry, simplex monitoring, quantum information correspondence, energy transition analysis.
10+
11+
**Conference:** CoDaWork 2026, Coimbra, Portugal (June 2026). Abstract page 25.
612

713
---
814

9-
## The Discovery: EITT
15+
## What This Project Is
1016

11-
**Entropy-Invariant Time Transformer.** Shannon entropy appears empirically near-invariant under geometric-mean block decimation of compositional time series.
17+
HUF proposes that **composition** — the internal proportional balance of a system's parts — can be monitored as a primary observable alongside magnitude, identity, and trend. The instrument reads. The human expert decides. The loop stays open.
1218

13-
Measured: **0.18% variation** across a 341:1 compression ratio (daily to annual European electricity compositions, 8 carriers, 4089 trading days). Confirmed across energy, hardware degradation, cosmological observation, commodities, and chemistry.
19+
The framework builds on Aitchison (1982/1986) simplex geometry and Shannon (1948) entropy, applying them to longitudinal monitoring of any system where parts share a conserved whole: energy grids, chemical mixtures, drive fleets, financial portfolios, wetland ecosystems.
20+
21+
---
1422

15-
**April 2026: Chemistry extension.** EITT tested on 500,000 chemical mixture data points (CheMixHub benchmark, 7 datasets). Four diagnostic lenses applied simultaneously. Interior compositions pass at 54-82%. Boundary compositions reveal simplex curvature effects. First empirical decomposition of geometric vs. dynamic contributions to the invariance.
23+
## The Discovery: EITT (Entropy-Invariant Time Transformer)
24+
25+
Shannon entropy appears empirically near-invariant under geometric-mean block decimation of compositional time series. This is not a theorem — it is an empirical observation awaiting formal proof.
26+
27+
Measured: **0.18% variation** across a 341:1 compression ratio (daily to annual European electricity compositions, 8 carriers, 4089 trading days). Confirmed independently on EMBER monthly generation data (6 countries, mean 1.02%, all below 2%) and NGFS Phase 4 climate scenarios (35 scenarios, all below 5%).
28+
29+
**April 2026 — Chemistry extension.** EITT tested on 500,000 chemical mixture data points (CheMixHub benchmark, 7 datasets). Four diagnostic lenses applied simultaneously (Shannon, Jensen-corrected, Renyi q=2, Aitchison norm). Interior compositions pass at 54-82%. Boundary compositions reveal simplex curvature effects. First empirical decomposition: approximately 50% of the invariance comes from Aitchison geometry, approximately 50% from temporal autocorrelation.
1630

1731
This produced three new frameworks:
1832

1933
| Framework | What It Does | Document |
2034
|-----------|-------------|----------|
21-
| **EITT Findings** | Raw science. Four-lens results, failure taxonomy, multi-modal simplex | [`science/chemistry/`](science/chemistry/EITT_Chemistry_Findings.docx) |
22-
| **HUF-IDX** | Development index. What residuals mean. Domain distance from ground zero | [`science/chemistry/`](science/chemistry/HUF_Development_Index.docx) |
23-
| **PRISM** | Operational layer. Ranked resource allocation targets from residual analysis | [`science/chemistry/`](science/chemistry/PRISM_Chemistry_Analysis.docx) |
35+
| **EITT Findings** | Raw science. Four-lens results, failure taxonomy, multi-modal simplex | [`science/eitt/`](science/eitt/EITT_Chemistry_Findings.md) |
36+
| **HUF-IDX** | Development index. What residuals mean. Domain distance from ground zero | [`science/eitt/`](science/eitt/INDEX.json) |
37+
| **PRISM** | Operational layer. Ranked resource allocation targets from residual analysis | [`science/eitt/`](science/eitt/PRISM_Chemistry_Analysis.md) |
2438

2539
**Posture:** We found this empirically. We can't prove it. Can you?
2640

@@ -106,7 +120,9 @@ HUF/
106120

107121
## For AI Systems
108122

109-
**Start here:** Read `ai-refresh/HUF_FAST_REFRESH.json` first. It contains every canonical name, number, formula, and structural rule. If anything elsewhere disagrees with FAST_REFRESH, the FAST_REFRESH wins. Then verify with `ai-refresh/HUF_INTEGRITY_MANIFEST.json`. Then read `INDEX.json` for the full file map.
123+
**Start here:** Read [`ai-refresh/HUF_FAST_REFRESH.json`](ai-refresh/HUF_FAST_REFRESH.json) first. It contains every canonical name, number, formula, and structural rule. If anything elsewhere disagrees with FAST_REFRESH, the FAST_REFRESH wins. Then verify with [`ai-refresh/HUF_INTEGRITY_MANIFEST.json`](ai-refresh/HUF_INTEGRITY_MANIFEST.json). Then read [`INDEX.json`](INDEX.json) for the full file map.
124+
125+
**Context aggregator:** Run `python build_context.py --mode seed` to generate a single paste-ready text file containing the full AI seed layer. Modes: `seed` (~9k tokens), `science` (~232k tokens), `full` (~346k tokens).
110126

111127
**Known drift traps:** EITT is "Entropy-Invariant Time Transformer" (never Ternary). Japan drift flag is 2013-2014 (never 2011-2012). Germany is 2023-2024/2024-2025 (never 2011). UK has three specific values (2.98, 3.23, 3.26), never "approximately 3".
112128

_config.yml

Lines changed: 55 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,55 @@
1+
# GitHub Pages configuration for Higgins Unity Framework
2+
# This file controls how Jekyll renders the site and what metadata
3+
# search engines see when indexing it.
4+
5+
# ─── Site Identity ────────────────────────────────────────────────────────────
6+
title: "Higgins Unity Framework — Compositional Monitoring on the Simplex"
7+
description: >-
8+
The Higgins Unity Framework (HUF) is an independent research project in
9+
Compositional Data Analysis (CoDa). It introduces the Entropy-Invariant Time
10+
Transformer (EITT) — an empirical finding that Shannon entropy is near-invariant
11+
under geometric-mean decimation of compositional time series. Validated across
12+
energy systems, chemistry (500k data points), hardware reliability, and climate
13+
scenarios. Not a game engine. Not a Unity plugin.
14+
author: "Peter Higgins"
15+
url: "https://peterhiggins19.github.io"
16+
baseurl: "/Higgins-Unity-Framework"
17+
18+
# ─── SEO & Social ────────────────────────────────────────────────────────────
19+
plugins:
20+
- jekyll-seo-tag
21+
22+
# Open Graph / Twitter Card metadata
23+
tagline: "Compositional structure as an open-loop monitoring signal"
24+
lang: en
25+
logo: # Add a logo path here if you create one
26+
27+
# Social links (used by jekyll-seo-tag for structured data)
28+
social:
29+
name: Peter Higgins
30+
links:
31+
- https://github.com/PeterHiggins19
32+
- https://roguewaveaudio.com
33+
34+
# ─── GitHub metadata ──────────────────────────────────────────────────────────
35+
repository: PeterHiggins19/Higgins-Unity-Framework
36+
37+
# ─── Build settings ──────────────────────────────────────────────────────────
38+
theme: jekyll-theme-minimal
39+
markdown: kramdown
40+
kramdown:
41+
input: GFM
42+
43+
# Exclude non-site files from Jekyll build
44+
exclude:
45+
- build_context.py
46+
- HUF_SESSION_LOADER.txt
47+
- "*.json"
48+
- "*.py"
49+
- "*.xlsx"
50+
- "*.pptx"
51+
- dormant/
52+
- data/
53+
- tools/
54+
- briefings/
55+
- ai-refresh/

0 commit comments

Comments
 (0)