|build| |docs| |madewithpython| |license|

SymDet

A python package to perform symmetry detection and generator extraction on raw data. Follows the paper by Sven Krippendorf and Marc Syvaeri on Detecting symmetries with neural networks <https://iopscience.iop.org/article/10.1088/2632-2153/abbd2d>_.

Notes

This project is under heavy development and is therefore not available on PyPi. I would not expect major API breaks but certainly addition of functionality.

Installation

Currently it is only possible to install SymDet from source and it will remain like this until it has been thoroughly tested on experimental data and a larger number of symmetry groups.

Install from source

pip installation

.. code-block:: bash

git clone https://github.com/SamTov/SymDet.git cd SymDet pip3 install . --user

conda installation

.. code-block:: bash

git clone https://github.com/SamTov/SymDet.git cd SymDet conda create -n SymDet python=3.8 conda activate SymDet pip3 install .

Documentation

There is a live version of the documentation hosted here <https://symdet.readthedocs.io/en/latest/>_. Alternatively you can build it from source using

.. code-block:: bash

cd Symdet/docs make html

You can then browse the documentation locally using your favourite browser.

Getting started

Because SymDet is not designed for single purpose, you will need to interface with different libraries and classes directly. This isn't as bad as it sounds and we have a number of tutorials to explain how this works. Broadly there are two modules relevant to most analysis, these are the analysis and generators modules.

• analysis: This module contains the necessary methods for analyzing raw data and extraction symmetry groups from it.
• generators: This module contains all of the modules necessary for extracting Lie group generators from the symmetry groups.

As a first step I would suggest looking at the examples <https://github.com/SamTov/SymDet/tree/main/examples>_ directory and following along with some tutorials.

Comments

This is a really young project and any comments or contributions would be welcome. If you see issues in the documentation (particularly if you're a mathematician) I would always welcome the feedback.

.. badges

.. |build| image:: https://github.com/SamTov/SymDet/actions/workflows/python-package.yml/badge.svg :alt: Build tests passing :target: https://github.com/SamTov/SymDet/blob/readme_badges/

.. |docs| image:: https://readthedocs.org/projects/symdet/badge/?version=latest&style=flat :alt: Build tests passing :target: https://symdet.readthedocs.io/en/latest/

.. |license| image:: https://img.shields.io/badge/License-EPLv2.0-purple.svg?style=flat :alt: Project license :target: https://www.gnu.org/licenses/quick-guide-gplv3.en.html

.. |madewithpython| image:: https://img.shields.io/badge/Made%20With-Python-blue.svg :alt: Made with python