Skip to content

Getting Started

Installation

chronos-lab can be used in several ways, depending on how you want to engage with the framework—from interactive prototyping to library-style usage to extending its internals. The options below reflect the most common and effective workflows.

1. Explore with JupyterLab + tutorials

If you want to understand how chronos-lab is structured and become productive quickly, the best entry point is JupyterLab together with the published tutorials.

All notebooks in the Tutorials section have fully rendered output, including charts, making them the easiest way to start exploring.

This path lets you see unified data access, pre-built Hamilton DAGs, ArcticDB-backed storage, and integrated visualization in action, without committing to a full local development setup.

Best for: first-time users, research prototyping, interactive analysis

Suggested steps:

  1. Start JupyterLab (locally, Google Colab, SageMaker, or similar)

  2. Clone the GitHub repository

  3. Work through the tutorial notebooks

git clone https://github.com/vitaliknet/chronos-lab.git
cd chronos-lab

This is the recommended starting point for most users.

2. Use chronos-lab as a Python library

If you already have an existing codebase and want to embed chronos-lab into your own pipelines, you can install it as a standard Python package.

This approach works well when you want to call chronos-lab DAGs, data access layers, or utilities from scripts, services, or notebooks that live outside this repository.

Best for: integration into existing projects, production pipelines, scheduled jobs

pip install chronos-lab

The base installation is intentionally minimal and includes only core dependencies. Most functionality is enabled via optional dependency groups:

pip install "chronos-lab[analysis,visualization,aws]"

Available extras include:

  • analysis – feature engineering, modeling, and calculation graph utilities

  • arcticdb – time-series storage with ArcticDB

  • aws – AWS integrations (S3, DynamoDB)

  • ib – Interactive Brokers market data (streaming and historical)

  • intrinio – Intrinio market data

  • visualization – matplotlib and mplfinance plotting

  • yfinance – Yahoo Finance market data

Extras can be combined as needed to match your environment.

3. Develop, extend, or compose your own DAGs

If you want to extend chronos-lab itself, modify existing pipelines, or build your own Hamilton DAGs using chronos-lab functions as composable building blocks, a full development setup is recommended.

This option gives you direct access to internal DAG definitions, shared utilities, and execution patterns used throughout the project.

Best for: contributors, advanced users, custom DAG authors

git clone https://github.com/vitaliknet/chronos-lab.git
cd chronos-lab

Install dependencies using uv, including all optional extras:

uv sync --all-extras

This installs chronos-lab in editable mode and mirrors the environment used by maintainers, making it easier to prototype new DAGs or adapt existing ones.

Python version support

chronos-lab requires Python 3.12 or newer. Python 3.13 is supported.

If you are unsure where to start, begin with JupyterLab and the tutorial notebooks, then move to a library or development setup as your needs become clearer.