Skip to main content

Deep neural network for solving continuous time economic models

Project description

Deep-MacroFin

Build Status PyPI version PyPI Downloads PyPI - License

Deep-MacroFin is a comprehensive deep-learning framework designed to solve equilibrium economic models in continuous time. The library leverages deep learning to alleviate curse of dimensionality.

Documentation: mkdocs

Start developing

Code

All the code are under deep_macrofin, and the tests are under tests

The project is now configured with uv for dependency management and packaging. To install the dependencies and run the code:

uv sync # this will automatically download suitable python versions, create virtual environment and install packages in pyproject.toml file

## You can now run the tests using the command:
pytest tests

To run the code and tests locally without poetry:

python -m venv venv
source venv/bin/activate # venv/Scripts/activate using Windows powershell
pip install -r requirements.txt
pip install -r requirements-dev.txt
pip install -r requirements-doc.txt

For easier testing, you can create a file in the root folder of the project, and import functions from deep_macrofin.

To properly run all tests in the tests/ folder

pip install -e .
pytest tests/

Examples

Various examples using the library, with comparisons to DeepXDE and PyMacroFin are included in examples

  • basic_examples: Solutions to basic ODEs/PDEs, diffusion equation, function approximation and systems of ODEs, with some comparisons to DeepXDE.
  • initial_examples: Initial scripts for testing deep neural networks for ODE/PDE solutions, and macromodels.
  • kan_examples: Solutions to basic ODEs, using KAN as approximators
  • macro_problems: Macroeconomic models in different dimensions.
  • paper_example: Examples in the paper, with PyMacroFin and deepXDE comparisons. Models and log files for reproducing paper results can be found in Google Drive. For a complete set of experiments, see this GitHub repo
  • pymacrofin_eg: Examples from PyMacroFin and proposition 2 from Brunnermeier and Sannikov (2014)

Note: PyMacroFin and DeepXDE are used as benchmarks in several examples, but the associated packages are not included in this repo's requirements.txt. To run the comparisons properly, please install their packages respectively.

Docs

The documentation site is based on mkdocs and mkdocs-mateiral.

Layouts

mkdocs.yml    # The configuration file.
docs/
    index.md  # The documentation homepage.
    ...       # Other markdown pages, images and other files.

To see the site locally, run the following command:

mkdocs serve

Cite Deep-MacroFin

If you use Deep-MacroFin for academic research, you are encouraged to cite the following paper:

@misc{wu2024deepmacrofin,
      title={Deep-MacroFin: Informed Equilibrium Neural Network for Continuous Time Economic Models}, 
      author={Yuntao Wu and Jiayuan Guo and Goutham Gopalakrishna and Zisis Poulos},
      year={2024},
      eprint={2408.10368},
      archivePrefix={arXiv},
      primaryClass={cs.LG},
      url={https://arxiv.org/abs/2408.10368}, 
}

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

deep_macrofin-0.5.1.tar.gz (32.0 MB view details)

Uploaded Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

deep_macrofin-0.5.1-py3-none-any.whl (144.5 kB view details)

Uploaded Python 3

File details

Details for the file deep_macrofin-0.5.1.tar.gz.

File metadata

  • Download URL: deep_macrofin-0.5.1.tar.gz
  • Upload date:
  • Size: 32.0 MB
  • Tags: Source
  • Uploaded using Trusted Publishing? Yes
  • Uploaded via: twine/6.1.0 CPython/3.13.12

File hashes

Hashes for deep_macrofin-0.5.1.tar.gz
Algorithm Hash digest
SHA256 35e7168989bce9c2273fef0a1af432b8bbe53cdc9da9c982fc9ea8082d7fe9b1
MD5 d3bf9b94cbb2ff04ea76adeed474e962
BLAKE2b-256 9474120abc7d53464cc771510e066a320588eafb468b10a77a19d8853670a129

See more details on using hashes here.

Provenance

The following attestation bundles were made for deep_macrofin-0.5.1.tar.gz:

Publisher: ci-build.yaml on rotmanfinhub/deep-macrofin

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

File details

Details for the file deep_macrofin-0.5.1-py3-none-any.whl.

File metadata

  • Download URL: deep_macrofin-0.5.1-py3-none-any.whl
  • Upload date:
  • Size: 144.5 kB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? Yes
  • Uploaded via: twine/6.1.0 CPython/3.13.12

File hashes

Hashes for deep_macrofin-0.5.1-py3-none-any.whl
Algorithm Hash digest
SHA256 7818854edbc47332948fc534ae8d4bd4c23534e5214cb180002915850f2ac6b2
MD5 ace87f224eb2de0e3ed5801f2b6bfce9
BLAKE2b-256 ac158b8651b7731f0fa6feb0e456a17a251ad4ffdb99a81dd00126c6a344f5d7

See more details on using hashes here.

Provenance

The following attestation bundles were made for deep_macrofin-0.5.1-py3-none-any.whl:

Publisher: ci-build.yaml on rotmanfinhub/deep-macrofin

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

Supported by

AWS Cloud computing and Security Sponsor Datadog Monitoring Depot Continuous Integration Fastly CDN Google Download Analytics Pingdom Monitoring Sentry Error logging StatusPage Status page