Pre-built GTSAM wheels for all platforms, published under a single package name so installation is the same everywhere.
There is no source code in this repo. It downloads official wheels from PyPI (gtsam-develop) and builds additional wheels with Nix for platforms/versions that are missing (e.g. Python 3.10 on macOS arm64). All wheels are renamed to gtsam-extended and uploaded to PyPI with a unified version.
pip install gtsam-extendedWorks as a drop-in replacement — just import gtsam as usual.
| macOS (arm64 + x86_64) | Linux x86_64 | Linux aarch64 (Jetson) | |
|---|---|---|---|
| Python 3.10 | yes (nix-built) | - | yes (cibuildwheel) |
| Python 3.11 | yes | yes | yes |
| Python 3.12 | yes | yes | yes |
| Python 3.13 | yes | yes | yes |
| Python 3.14 | yes | yes | yes |
See PUBLISHING.md for details.
Important Note
As of Dec 2021, the develop branch is officially in "Pre 4.2" mode. A great new feature we will be adding in 4.2 is hybrid inference a la DCSLAM (Kevin Doherty et al) and we envision several API-breaking changes will happen in the discrete folder.
In addition, features deprecated in 4.1 will be removed. Please use the last 4.1.1 release if you need those features. However, most (not all, unfortunately) are easily converted and can be tracked down (in 4.1.1) by disabling the cmake flag GTSAM_ALLOW_DEPRECATED_SINCE_V42.
GTSAM is a C++ library that implements smoothing and mapping (SAM) in robotics and vision, using Factor Graphs and Bayes Networks as the underlying computing paradigm rather than sparse matrices.
The current support matrix is:
| Platform | Compiler | Build Status |
|---|---|---|
| Ubuntu 18.04 | gcc/clang | |
| macOS | clang | |
| Windows | MSVC |
On top of the C++ library, GTSAM includes wrappers for MATLAB & Python.
mkdir build
cd build
cmake ..
make check (optional, runs unit tests)
make installPrerequisites: Boost >= 1.65, CMake >= 3.0, a modern compiler (at least gcc 4.7.3 on Linux). Optional: Intel TBB, Intel MKL.
Support for MATLAB and Python wrappers is provided.
The recommended citation uses the BibTeX entry for borglab/gtsam (Frank Dellaert and GTSAM Contributors, version 4.2a8, 2022). Additional citations are provided for the "Factor Graphs for Robot Perception" book and the IMU preintegration scheme.
Includes a state-of-the-art IMU handling scheme based on work by Lupton/Sukkarieh (2012) and Forster/Carlone/Dellaert/Scaramuzza (2015), with an efficient implementation integrating on the NavState tangent space.
- GTSAM users Google group
- Open source under the BSD license
- Developed in the lab of Frank Dellaert at Georgia Institute of Technology