{"id":13449700,"url":"https://github.com/jankrepl/deepdow","last_synced_at":"2026-01-17T16:58:51.441Z","repository":{"id":40001743,"uuid":"237742797","full_name":"jankrepl/deepdow","owner":"jankrepl","description":"Portfolio optimization with deep learning.","archived":false,"fork":false,"pushed_at":"2024-01-24T15:56:49.000Z","size":2354,"stargazers_count":985,"open_issues_count":27,"forks_count":142,"subscribers_count":26,"default_branch":"master","last_synced_at":"2025-03-05T23:48:56.976Z","etag":null,"topics":["allocation","convex-optimization","deep-learning","finance","machine-learning","markowitz","portfolio-optimization","pytorch","stock-price-prediction","timeseries","trading","wealth-management"],"latest_commit_sha":null,"homepage":"https://deepdow.readthedocs.io","language":"Python","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"apache-2.0","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/jankrepl.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":"LICENSE","code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null,"roadmap":null,"authors":null,"dei":null}},"created_at":"2020-02-02T08:46:33.000Z","updated_at":"2025-03-05T11:38:26.000Z","dependencies_parsed_at":"2024-01-07T10:50:59.543Z","dependency_job_id":"22d1e301-cc49-4b99-96bb-4758a3b82d8c","html_url":"https://github.com/jankrepl/deepdow","commit_stats":null,"previous_names":[],"tags_count":6,"template":false,"template_full_name":null,"repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/jankrepl%2Fdeepdow","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/jankrepl%2Fdeepdow/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/jankrepl%2Fdeepdow/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/jankrepl%2Fdeepdow/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/jankrepl","download_url":"https://codeload.github.com/jankrepl/deepdow/tar.gz/refs/heads/master","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":245029292,"owners_count":20549682,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2022-07-04T15:15:14.044Z","host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["allocation","convex-optimization","deep-learning","finance","machine-learning","markowitz","portfolio-optimization","pytorch","stock-price-prediction","timeseries","trading","wealth-management"],"created_at":"2024-07-31T06:00:51.753Z","updated_at":"2026-01-17T16:58:51.423Z","avatar_url":"https://github.com/jankrepl.png","language":"Python","funding_links":[],"categories":["Analytics","Strategies \u0026 Research","Python","Analytic tools"],"sub_categories":["Optimization","Portfolio Management","Trading \u0026 Backtesting","交易与回测"],"readme":"![final](https://user-images.githubusercontent.com/18519371/79003829-afca6380-7b53-11ea-8322-f05577536957.png)\n\n[![codecov](https://codecov.io/gh/jankrepl/deepdow/branch/master/graph/badge.svg)](https://codecov.io/gh/jankrepl/deepdow)\n[![Documentation Status](https://readthedocs.org/projects/deepdow/badge/?version=latest)](https://deepdow.readthedocs.io/en/latest/?badge=latest)\n[![PyPI version](https://badge.fury.io/py/deepdow.svg)](https://badge.fury.io/py/deepdow)\n[![DOI](https://zenodo.org/badge/237742797.svg)](https://zenodo.org/badge/latestdoi/237742797)\n\n`deepdow` (read as \"wow\") is a Python package connecting portfolio optimization and deep learning. Its goal is to\nfacilitate research of networks that perform weight allocation in **one forward pass**.\n\n\n# Installation\n```bash\npip install deepdow\n```\n# Resources\n- [**Getting started**](https://deepdow.readthedocs.io/en/latest/auto_examples/end_to_end/getting_started.html)\n- [**Detailed documentation**](https://deepdow.readthedocs.io/en/latest)\n- [**More examples**](https://deepdow.readthedocs.io/en/latest/auto_examples/index.html)\n\n# Description\n`deepdow` attempts to **merge** two very common steps in portfolio optimization\n1. Forecasting of future evolution of the market (LSTM, GARCH,...)\n2. Optimization problem design and solution (convex optimization, ...)\n\nIt does so by constructing a pipeline of layers. The last layer performs the allocation and all the previous ones serve\nas feature extractors. The overall network is **fully differentiable** and one can optimize its parameters by gradient\ndescent algorithms.\n\n# `deepdow` is not ...\n- focused on active trading strategies, it only finds allocations to be held over some horizon (**buy and hold**)\n    - one implication is that transaction costs associated with frequent, short-term trades, will not be a primary concern \n- a reinforcement learning framework, however, one might easily reuse `deepdow` layers in other deep learning applications\n- a single algorithm, instead, it is a framework that allows for easy experimentation with powerful building blocks\n\n\n# Some features\n- all layers built on `torch` and fully differentiable\n- integrates differentiable convex optimization (`cvxpylayers`)\n- implements clustering based portfolio allocation algorithms\n- multiple dataloading strategies (`RigidDataLoader`, `FlexibleDataLoader`)\n- integration with `mlflow` and `tensorboard` via callbacks\n- provides variety of losses like sharpe ratio, maximum drawdown, ...\n- simple to extend and customize\n- CPU and GPU support\n\n# Citing\nIf you use `deepdow` (including ideas proposed in the documentation, examples and tests) in your research please **make sure to cite it**.\nTo obtain all the necessary citing information, click on the **DOI badge** at the beginning of this README and you will be automatically redirected to an external website.\nNote that we are currently using [Zenodo](https://zenodo.org/).\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fjankrepl%2Fdeepdow","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fjankrepl%2Fdeepdow","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fjankrepl%2Fdeepdow/lists"}