SnapPy

What is SnapPy?

SnapPy is a program for studying the topology and geometry of 3-manifolds, with a focus on hyperbolic structures. It runs on Mac OS X, Linux, and Windows, and combines a link editor and 3D-graphics for Dirichlet domains and cusp neighborhoods with a powerful command-line interface based on the Python programming language. You can see it in action, learn how to install it, and watch the tutorial.

News

Version 3.3 (January 2026):
- Link now supports band moves and can search for ribbon disks and ribbon concordances. See ribbon_concordant_links and add_band. From [DG].
- New census RibbonLinks. From [DG].
- Additional slice obstructions added to slice_obstruction_HKL. From [DG].
- The Fox-Milnor slice obstruction is now available as fox_milnor_test. From [DG].
- The census OrientableCuspedCensus has been extended to 10 ideal tetrahedra by [Li], adding 150,000 new manifolds.
- Margulis number can now be computed with margulis.
- The upper bounds on the bridge number of a link from [BKVV2020] and [BKP2025] are available as bridge_upper_bound.
- triangulation_isosig now preserves the orientation of the core curve rather than the Dehn-filling curve - thus, preserving the (unoriented) spun-triangulation structure. This only changes the result when ignore_curves=True (and ignore_orientation and ignore_filling_orientations have their default values).
- Fixing a bug in length_spectrum_alt and length_spectrum_alt_gen. The estimate for the upper bound of the global spine radius was not correct in all cases. Note that we expect this bug to be very hard to hit and do not have a single example where the result of these methods was wrong.
- length_spectrum_alt can be given count and max_len at the same time. In that case, it stops when enough geodesics have been listed to ensure that they include the count shortest geodesics or that they include all geodesics shorter than max_len.
- Acceleration to length_spectrum_alt when the next geodesic has large gap to the last geodesic in the result. It uses the new include_intermediates=True flag of length_spectrum_alt_gen.
- Computing maximal cusp areas in an unbiased way uses a simpler algorithm which also returns tighter intervals for verified computations. This affects cusp_areas and related methods.
- Support for Python 3.14 and SageMath 10.8.
Version 3.2 (January 2025):
- An alternative implementation of the length spectrum. It supports verified computations and is significantly faster in some cases. See length_spectrum_alt and length_spectrum_alt_gen.
- isometry_signature now also works for closed manifolds.
- A paper plane or eye ball in the inside_view showing the position and bearing of the user in the hyperbolic manifold. By default, we now show the paper plane instead of the edges of the triangulation so that the view is intrinsic to the manifold. The geodesics window now also features a button to put the camera onto a geodesic. Here are some examples:
- A faster and more robust algorithm to the compute maximal cusp area matrix. The new algorithm is now the default for cusp_area_matrix(), cusp_areas(), short_slopes() and cusp_translations().
- New options ignore_curves and ignore_filling_orientations for triangulation_isosig(). Also fixing a subtle bug where the filling coefficients returned by triangulation_isosig were not canonical when ignore_curve_orientations = True.
- canonical_retriangulation() and isometry_signature() fail with exceptions rather than silently returning None. In particular, it now safe to compare isometry signatures (without further checks) to determine whether M and N are isometric hyperbolic manifolds:
```
>>> M.isometry_signature(verified=True) == N.isometry_signature(verified=True)
```
- Bug fix to slice_obstruction_HKL: earlier versions incorrectly allowed looking at mod 2 homology of the branched cover, where the theory does not directly apply.
- New self-contained SnapPy application for Linux.
- Support for Python 3.13 and SageMath 10.5.
Complete version history.

Documentation

Credits

Written by Marc Culler, Nathan Dunfield, and Matthias Goerner using the SnapPea kernel written by Jeff Weeks, with contributions from many others. If you use SnapPy in your work, please cite it as described here. If you encounter problems with SnapPy, please report them.

Released under the terms of the GNU General Public License, version 2 or later.

The development of SnapPy was partially supported by grants from the National Science Foundation, including DMS-0707136, DMS-0906155, DMS-1105476, DMS-1510204, DMS-1811156, and DMS-2303572, and the Institute for Computational and Experimental Research in Mathematics. Any opinions, findings, and conclusions or recommendations expressed on this site are those of the authors and do not necessarily reflect the views of the National Science Foundation.