Human Fictions in Monetary Policy

The project enables economists to simulate how human frictions influence responses to monetary policy. The Gym-style environment reflects classic macro-economic model with human agents representing households. The agents can learn to respond optimally with simple heuristics or deep reinforcement learning (e.g. with rllib).

Prerequisites

Before you begin, ensure you have met the following requirements:

• You have a Linux/Mac machine.
• You have docker installed.

Development

First usage

To start, run this line in human-friction dir:

docker build -t human-friction:1.0 .

It will build docker image, used for running other commands. You can execute scripts using Docker. Docker must be build to reflect changes.

docker run human-friction:1.0

Daily development

To start development, edit files in your editor. You can execute scripts using Docker. Docker must be build to reflect changes.

docker run human-friction:1.0

Alternatively: Virtualenv instead of Docker

python3 -m venv venv
. venv/bin/activate
pip3 install --upgrade pip setuptools wheel
pip3 install -e .
pip3 install -r requirements.txt

Then you can run a training script from command line, e.g.:

python3 train.py

Pre-commit

Make sure that code style follows our guidlines by installing pre-commit: Using pip:

pip install pre-commit

Using homebrew:

brew install pre-commit

Then run

pre-commit install

to set up the git hook scripts.

Packages management

If you need an additional package in docker, add it to requirements.in and run script ./scripts/run-refresh-requirements. It will refresh requirements.txt file, based on content of requirements.in. When finished, you have to rebuild your docker image to include new packages using docker build -t human-friction:1.0 . command. You also have to do it if someone else updated requirements.txt and you pulled it from remote repository.

Using human-friction

To use human-friction, train and evaluate the agents, an example is provided in train.py and evaluate.py.

Contributing to Human Frictions

To contribute to human-frictions, follow these steps:

1. Clone this repository.
2. Create a branch: git checkout -b <branch_name>.
3. Make your changes and commit them: git commit -m '<commit_message>'
4. Push to the original branch: git push origin
5. Create the pull request.

Alternatively see the GitHub documentation on creating a pull request.

Testing

We are using pytest`for tests. You can run them via:

pytest

Contributors

Thanks to the following people who have contributed to this project:

• @mesjou 💻
• @annaalm 💻

Contact

If you want to contact me you can reach me at matthias.hettich@tu-berlin.de.

License

This project uses the following license: MIT License.