cmu-phil/tetrad

Repository for the Tetrad Project, www.phil.cmu.edu/tetrad.

What it solves

Tetrad is a software suite designed for causal discovery and analysis, providing tools to identify causal relationships between variables from data.

How it works

The project provides a core library of model code (tetrad-lib) and a graphical user interface (tetrad-gui) for interacting with the data. It is available as a standalone application, a Python wrapper (py-tetrad) for integration into Python workflows, and a limited R integration (rpy-tetrad), as well as a command-line tool called Causal Command.

Who it’s for

Researchers and programmers who need to perform causal searches and analysis using Java, Python, or R.

Highlights

  • Standalone GUI application for causal analysis.
  • Python and R integrations for programmatic access to Tetrad functionality.
  • Command-line tool for running arbitrary Tetrad searches.
  • Built using Java and Maven for easy compilation and packaging.

Related

  • Project
    erdogant/bnlearn

    bnlearn is a Python library for learning, parameterising, and querying Bayesian networks. It supports structure discovery (various scores and search methods), CPT estimation, causal inference (do‑calculus), synthetic data generation, and visualisation, all wrapped in a simple API.

  • Project
    Kanaries/Rath

    An open-source augmented analytics tool that automates exploratory data analysis by discovering patterns and generating multi-dimensional visualizations.

  • Project
    scikit-tda/scikit-tda

    A collection of Topological Data Analysis (TDA) Python libraries designed to make topological data tools accessible to non-topologists.

  • Project
    pgmpy/pgmpy

    A Python library for causal and probabilistic reasoning using graphical models, providing tools for structure learning, parameter estimation, and causal inference.

  • Dispatch
    Claude Developer Platform Advanced Tool Use

    Anthropic has introduced Tool Search Tool, Programmatic Tool Calling, and Tool Use Examples to enable Claude agents to manage thousands of tools and large datasets without overloading the context window.