Dynamic Graphical Variable Selection (DGVS)

Dynamic Graphical Variable Selection (DGVS) is an algorithm for estimating dynamic graphical structures in time-series data. This repository contains the main estimation function, simulation scripts, and example results as presented in the paper:

Dynamic Graphical Models with Variable Selection for Effective Connectivity Estimation
DOI: 10.1214/23-BA1377

📦 Technologies

• R
• C++ (via Rcpp)

🚀 Installation and Usage

1. Download the files DGVScpp.cpp and Main_DGVS.r.

2. Open Main_DGVS.r in R.

3. Ensure the path to the DGVScpp.cpp file is correctly specified in the script (e.g., use setwd() or relative paths as needed).

4. Run the entire Main_DGVS.r script. This will compile and execute the DGVS algorithm and display the results.

✉️ Contributions

This project is not currently open to public contributions via pull requests. However, suggestions, questions, or collaboration inquiries are welcome by email:

📧 rebecca@dme.ufrj.br

📄 License

No formal license is defined for this repository. Please contact the author before reuse or distribution of the code or results.

📚 Citation

If you use this code in your work, please cite the original paper:

Souza, R. O., Rodrigues, J., Lopes, H. F., & Macchiavelli, R. E. (2024). Dynamic Graphical Models with Variable Selection for Effective Connectivity Estimation. Bayesian Analysis, 19(4), 1001–1032. https://doi.org/10.1214/23-BA1377