Py-DYNAMIN

Python DYNamic Agent-based MINskyan (Py-DYNAMIN) model.

The Py-DYNAMIN model is a reusable agent-based macroeconomic modelling framework implemented in Python. It allows you to run baseline economic scenarios and post-growth scenario analysis for research and reproducibility purposes, the PG-DYNAMIN model. The model requires PostgreSQL for database operations and standard Python scientific libraries for analysis.

Features

• Agent-based macroeconomic modeling
• Baseline scenario and post-growth scenarios
• PostgreSQL integration for data storage
• Analysis tools for macro, micro, and stylised facts
• Example scripts for academic reproducibility (PG-DYNAMIN scenario analysis)

Installation

You can install the baseline model Py-DYNAMIN via pip from PyPI using:

pip install py-dynamin

Or, clone the repository for full access to the code, examples (PG-DYNAMIN) and analysis scripts:

git clone https://github.com/DylanTerryDoyle/py-dynamin.git

Dependencies:

• Python >= 3.10

• Core: numpy, scikit-learn, psycopg2-binary, pyyaml, tqdm

• Optional for analysis: pandas, matplotlib, scipy, powerlaw, statsmodels

• Database: PostgreSQL must be installed and running, see: https://www.postgresql.org/ for installation details.

Project Structure

py-dynamin/
├── LICENSE
├── README.md
├── pyproject.toml
├── src
│   └── dynamin
│        ├── __init__.py
│        ├── agents
│        │   ├── abtract_firm.py
│        │   ├── banks.py
│        │   ├── capital_firm.py
│        │   ├── consumption_firm.py
│        │   └── household.py
│        ├── config
│        │   ├── database.yaml
│        │   └── parameters.yaml
│        ├── baseline.py
│        ├── data_collector.py
│        ├── model.py
│        ├── simulator.py
│        └── utils.py
├── examples
│   └── post_growth_scenarios
│       ├── config
│       │   ├── database.yaml
│       │   ├── parameters.yaml
│       │   └── pg_scenarios.yaml
│       └── pg_scenarios.py
├── analysis
│   ├── __init__.py
│   ├── empirical_data
│   │   ├── CE16OV.csv
│   │   ├── GDPC1.csv
│   │   ├── GPDIC1.csv
│   │   ├── PCECC96.csv
│   │   ├── TBSDODNS.csv
│   │   └── UNRATE.csv
│   ├── esl.py
│   ├── macro.py
│   ├── micro.py
│   ├── stylised_facts.py
│   └── utils.py
└── docs
    └── DyNAMIN Model Description.pdf

• src/dynamin: core library for the baseline model.
• src/dynamin/config: baseline parameters and database configuration (database.yaml), update user/password to match your PostgreSQL setup.
• examples/post_growth_scenarios: parameters and scripts for running post-growth scenario analysis.
• analysis: scripts for running batch analyses of model results.
• analysis/empirical_data: empirical time-series for stylised_facts.py analysis.

Quick Start

1. Ensure PostgreSQL is installed and running. Update the database.yaml user/password fields to match your database credentials.
2. Install Py-DYNAMIN via pip:

pip install py-dynamin

3. Run baseline scenario (installed in package):

python -m dynamin.baseline

4. Run post-growth scenarios (from repository)

To run the post-growth scenario examples, clone the repository to access the examples folder:

git clone https://github.com/DylanTerryDoyle/py-dynamin.git
cd py-dynamin
python examples/post_growth_scenarios/pg_scenarios.py

5. Run analysis (from repository)

After running the simulations, you can analyze results using the analysis scripts:

git clone https://github.com/DylanTerryDoyle/py-dynamin.git
cd py-dynamin
python analysis/stylised_facts.py
python analysis/macro.py
python analysis/micro.py

Results will be printed to the terminal and figures can be found in the analysis figures folders

Contributing

Contributions are welcome. Please fork the repository, make changes, and submit a pull request. Ensure that all new code includes tests where applicable.

License

This project is licensed under the GNU Affero General Public License v3.0. See LICENSE for details.

Contact

Dylan Terry-Doyle
Email: d.c.terry-doyle@sussex.ac.uk