For a hands-on mathematical introduction, start with Factor Graphs and GTSAM: A Hands-on Introduction.

For a broader treatment of factor graphs in robotics, see Factor graphs for robot perception by Frank Dellaert and Michael Kaess.

The current stable release is GTSAM 4.3.0. The develop branch contains changes intended for the next release and may include API changes.

Documentation

Choose the documentation that matches what you need:

• User guide: installation, core concepts, tutorials, and module guides
• Example notebooks: worked examples, many with Colab links; module API notebooks are also available through the user guide. Requirements vary by example.
• C++ API reference: generated reference documentation for GTSAM 4.3.0

Language Wrappers

GTSAM also ships with Python and MATLAB wrappers. For build and packaging details, see the upstream wrapper READMEs:

• Python wrapper README
• MATLAB wrapper README

For CUDA from Python, build GTSAM and its Python wrapper on a CUDA-equipped machine. The standard GTSAM 4.3.0 Python wheels do not include the experimental CUDA bindings.

Notes on GTSAM

• GTSAM Concepts: manifolds, groups, Lie groups, vector spaces, and traits
• The Preintegrated IMU Factor: overview of GTSAM’s IMU handling scheme
• Migrating from GTSAM 3: highlights of the major GTSAM 4 changes
• Contributing to GTSAM: contribution workflow, review expectations, and coding conventions

Additional Information

There is a GTSAM users Google group for general discussion.