Fanti Mechanical Dating — Reproducible Evaluation Codebase

Python code, data and figures supporting the article
“Critical Assessment of Mechanical Dating Methods for Ancient Flax Fibres: Viscoelastic Modelling Versus Multilinear Regression.”

📝 Associated Article

Title : Critical Assessment of Mechanical Dating Methods for Ancient Flax Fibres: Viscoelastic Modelling Versus Multilinear Regression
Author : Michel Bakhtaoui

Concept DOI : 10.5281/zenodo.15098838 (always resolves to the latest release)

🔍 Project Overview

This repository contains:

• Statistical re-analysis of Fanti et al. (2015) mechanical dating data.
  
• Physics-based viscoelastic modelling (multi-branch Maxwell).
  
• Bayesian inference and Monte-Carlo uncertainty propagation.
  
• Reproducible figures and scripts used in the manuscript.

All code is MIT-licensed and fully reproducible.

🗂 Repository Structure

text . ├── CITATION.cff ├── data │ └── fanti_mechanical_data.csv ├── LICENSE ├── README.md ├── requirements.txt └── scripts ├── modeling │ ├── fanti_comparison.py │ ├── figure1.png │ ├── figure2.png │ ├── figure4.png │ ├── figure6.png │ ├── figure_comparison_fanti.png │ └── modelisation.py └── statistical-analysis ├── analysis.py ├── fanti_crossed_analysis.py ├── fanti_cross_validation.py ├── fanti_experiment.py ├── fanti_multiple_regression.py ├── fanti_power_analysis.py ├── fanti_uncertainty_propagation.py ├── figure2.png ├── figure3.png └── figure5.png

📊 Figures Generated

| Figure | Description | Script | |--------|----------------------------------------------------|-------------------------------------| | 1 | Age vs mechanical resistance (Maxwell fit) | modelisation.py | | 2 | Regression model performance (fit & cross-val) | fanti_multiple_regression.py | | 3 | Monte-Carlo R² distribution | analysis.py | | 4 | Posterior age distribution (Bayesian) | modelisation.py | | 5 | Variance Inflation Factor (VIF) | fanti_multiple_regression.py | | 6 | MCMC trace & convergence diagnostics | modelisation.py | | 7 | Fanti data vs viscoelastic model comparison | fanti_comparison.py |

🚀 Quick Start

1 • Installation

bash git clone https://github.com/mab-plus/fanti-mechanical-dating-analysis.git cd fanti-mechanical-dating-analysis python -m pip install -r requirements.txt

2 • Run full viscoelastic model

bash cd scripts/modeling python modelisation.py # generates Figures 1, 4, 6

3 • Reproduce Fanti cross-validation

bash cd ../statistical-analysis python fanti_cross_validation.py # generates cross-val metrics & Figure 2

4 • Physics vs empirical comparison

bash cd ../modeling python fanti_comparison.py # generates Figure 7

All generated figures are saved in the corresponding script folders.

📚 Data Sources

| Dataset | Location | Licence | |---------|----------|---------| | Mechanical data (Fanti et al. 2015, 9 samples) | data/fanti_mechanical_data.csv | CC-BY | | Example Maxwell parameters | in-code (modelisation.py) | MIT |

🤝 How to Cite

Bakhtaoui M. (2025). Fanti Mechanical Dating Analysis. Zenodo.
https://doi.org/10.5281/zenodo.15098838

📑 Licence

This project is released under the MIT License (see LICENSE).

🙏 Acknowledgements

Thanks to Dr Alain Bourmaud for providing nano-indentation data, and to the open-source community (NumPy, SciPy, Matplotlib, etc.).