JOR-Framework-PyMC

PyMC implementation of the James Orion Report (JOR) Framework for UAP case analysis.
This repository contains scripts to calculate SOP, NHP, and posterior probabilities for JOR Framework v3 cases using Bayesian methods.

---

Framework Intent

The JOR Framework is a probabilistic evidence model designed to constrain interpretation of UAP cases, not to assume or promote specific conclusions.

SOP (Solid Object Probability) establishes the evidentiary foundation, and NHP (Non-Human Probability) is conditionally dependent on that foundation. This structure prevents overclassification and limits non-human attribution in the absence of sufficient supporting evidence.

The model is intended for analytical use in evaluating case data under uncertainty, with an emphasis on evidence weighting and controlled inference.

Workflow Overview

1. Run jor_fusion.py → generates jor_scores.csv with SOP/NHP for each case
2. Run jor_pymc_runner.py → reads jor_scores.csv, performs Bayesian analysis, updates it with posterior means and distributions

jor_fusion.py
      ↓
   jor_scores.csv (initial SOP/NHP)
      ↓
jor_pymc_runner.py
      ↓
   jor_scores.csv (updated with posterior means and distributions)

---

Repository Contents

• jor_fusion.py – Generates jor_scores.csv with SOP/NHP scores
• jor_pymc.py – Main PyMC Bayesian model for JOR cases
• jor_pymc_runner.py – Reads jor_scores.csv and updates it with posterior means and distributions

---

Model Calibration & Precision

The Bayesian implementation of JOR uses informative Beta priors to incorporate expert evidentiary input.

• Sigma (σ) Tuning: Input sigmas for Credibility (C), Evidence (E), and Physicality (P) are calibrated at 0.02.
• Analytical Result: This calibration stabilizes the Average Uncertainty (95% credible interval width) at ~11.2%, yielding a 22% increase in precision over the baseline model while maintaining realistic margins for historical cases.
• Inference Constraints: The model enforces a baseline skepticism floor of 0.20 and an approximate posterior ceiling of 0.50, ensuring that Non-Human Probability (NHP) remains strictly grounded in the evidentiary foundation provided by SOP.

---

Requirements

• Python 3.10+
• PyMC
• NumPy
• Pandas

Install dependencies with pip:

pip install pymc numpy pandas

---

How to Run

1. Generate initial case scores:

python jor_fusion.py

2. Run Bayesian analysis:

python jor_pymc_runner.py

3. Outputs:

• jor_scores.csv updated with:
  • Original SOP and NHP values
  • Posterior means
  • Posterior distributions

4. Optional: Call jor_pymc.py directly for integration or custom analysis.

---

Notes

• NHP is conditionally dependent on SOP according to JOR Framework logic
• This repo is a self-contained PyMC workflow for JOR case analysis
• Future updates may include additional automation for case imports and helper scripts

---

License

This project is licensed under the MIT License

Citation / DOI

If you use this repository or its datasets in your research, please cite the Zenodo release:

James Orion Report (JOR) Framework PyMC Dataset, Zenodo, DOI: 10.5281/zenodo.18157347