This package provides wrappers for using various persistent homology and other Topological Data Analysis software packages through python.
The code is an compilation of work done by Elizabeth Munch along with her students and collaborators. People who have contributed to teaspoon include:
- Firas Khasawneh
- Brian Bollen
Please note that this code is still pre-alpha, so many things are not fully up and running yet.
Bugs reports and feature reqests can be posted on the gitlab issues page.
In order to use all wrappers in teaspoon.TDA, the following need to be installed so they can be run command line.
- Ripser. Code by Ulrich Bauer for computing persistent homology of a point cloud or distance matrix.
- Perseus. Code by Vidit Nanda for computing persistent homology of point clouds, cubical complexes, and distance matrices.
- Hera. Code by Michael Kerber, Dmitriy Morozov, and Arnur Nigmetov for computing bottleneck and Wasserstein distances.
Warning: the teaspoon installation will not install nor check for the install of these packages. In particular, perseus should be installed so that typing
perseus in a terminal runs it, even though the default perseus installation for some operating systems ends up with an executable with a different name.
This code is still pre-alpha. In particular, using the pip install seems to be finicky at best.
If you are having installation issues, please make a note of what you've done, including copying error message outputs, as a comment into the installation issue on gitlab so we can start figuring out what is up with this system.
Installing using pip:
So far, the most success has been had with pip installation. Run the following commands to clone and install teaspoon. In particular, the
pip install . command needs to be run in the folder with the
git clone https://gitlab.msu.edu/TSAwithTDA/teaspoon cd teaspoon pip install .
If you already have teaspoon installed but want to update to a newer version, run the following from the teaspoon folder.
pip install -U .
Installing using python:
According to the internet, the pip version of install appears to be better. However, if you don't use pip, another option is to cd into the teaspoon directory and run:
python setup.py install
Alternatively, just add the folder to the python path.
If the above aren't working, add the following line to your
Documentation is done using doxygen. The most recent version of the teaspoon documentation is hosted here, but can also be found locally in the doc folder. Further info on documentation can be found in the contributing page.
See the contributing page for more information on workflows.
Liz Munch: firstname.lastname@example.org.