Bayesian MCMC in Rate State Models

Directory Structure

Bayesian_MCMC_Deep-Learning/
│
├── .github/ - Contains GitHub Actions workflow files.
│   └── workflows/
│       └── python-app.yml - Configuration for Python application workflow.
│
├── .vscode/ - Contains settings for Visual Studio Code.
│   ├── launch.json - Debugging configurations.
│   ├── settings.json - VSCode settings.
│   └── tasks.json - Task configurations for VSCode.
│
├── source_c++/ - C++ source files for MCMC and deep learning inference.
│   ├── MCMC.cpp - Markov Chain Monte Carlo implementation.
│   ├── MCMC.h
│   ├── RateStateModel.cpp - Rate state models for simulations.
│   ├── RateStateModel.h
│   └── dl_inference.cpp - Deep learning inference related scripts.
│
├── source_python/ - Python scripts for MCMC, deep learning, and utilities.
│   ├── MCMC.py - Markov Chain Monte Carlo implementation in Python.
│   ├── RSF.py - Related to rate state functions or models.
│   ├── RateStateModel.py - Python implementation of rate state models.
│   ├── __init__.py - Initializes the Python package.
│   ├── __pycache__/ - Compiled Python files for faster loading.
│   ├── dl_inference.py - Scripts for deep learning inference.
│   ├── lstm/ - Contains LSTM related models and utilities.
│   │   ├── lstm_encoder_decoder.py
│   │   └── utils.py
│   ├── tests/ - Test scripts for MCMC and other functionalities.
│   │   ├── __init__.py
│   │   ├── test_mcmc.py
│   │   └── test_rsf.py
│   └── utils/ - Utility scripts for operations like JSON and MySQL interactions.
│       ├── json_save_load.py
│       └── mysql_save_load.py
│
├── .DS_Store - A file used by macOS to store custom attributes of a folder.
├── README.md - The repository's readme file with an overview and instructions.
├── __init__.py - An initialization script for Python packages.
└── requirements.txt - Lists the Python dependencies for the project.

Introduction

Adaptive Metropolis Algorithm in a nutshell

Running the Source code Directly

Python Version

1. Install dependencies.
2. Run Script: python3 -m source_python.dl_inference

C++ Version

1. Install dependencies (boost, gsl, eigen)
2. Navigate to source_c++ folder
3. Compile: g++ -std=c++17 -o executable dl_inference.cpp RateStateModel.cpp MCMC.cpp [...]
4. Run executable

Features

• Rate State Model Initialization with customizable parameters.
• Time Series Generation and Bayesian Inference using MCMC.
• Visualization with matplotlib.

Paper Summary

Based on "Arriving at estimates of a rate and state fault friction model parameter using Bayesian inference and Markov chain Monte Carlo", https://www.sciencedirect.com/science/article/pii/S266654412200003X

Contact

For support or collaboration: dana.spk5@gmail.com

---

Copyright © [Saumik Dana]