ITMO-NSS-team / torch_DE_solver

Combines power of torch, numerical methods to conquer and solve ALL {O,P}DEs
https://torch-de-solver.readthedocs.io
BSD 3-Clause "New" or "Revised" License
59 stars 9 forks source link
differential-equations equation-solver numerical-methods

TEDEouS - Torch Exhaustive Differential Equation Solver

The purpose of the project

  1. Make equation discovery more transparent and illustrative
  2. Combine power of pytorch, numerical methods and math overall to conquer and solve ALL XDEs(X={O,P}). There are some examples to provide a little insight to an operator form

Table of Contents

Core features

Installation

TEDEouS can be installed with pip::

$ git clone https://github.com/ITMO-NSS-team/torch_DE_solver.git $ cd torch_DE_solver $ pip install -r requirements.txt

Examples

After the TEDEouS is installed the user may refer to various examples that are in examples forlder. ::

$ cd examples

Every example is designed such that the boxplots of the launches are commented and the preliminary results are not shown, but stored in separate folders.

::

$ python example_ODE_Legendre.py

or ::

$ python example_ODE_Legendre_autograd.py

::

$ python example_Painleve_I.py

::

$ python example_wave_paper_autograd.py

::

$ python example_wave_physics.py

::

$ python example_heat.py

::

$ python example_KdV.py

::

$ python example_KdV_periodic.py

::

$ python example_Burgers_paper.py

Project Structure

Stable version is located in the master branch.

Documentation

https://torch-de-solver.readthedocs.io/en/docs/index.html

Getting started

Schroedinger equation example step-by-step https://torch-de-solver.readthedocs.io/en/docs/tedeous/examples/schrodinger.html

License

TEDEouS is distributed under BSD-3 licence found in LICENCE file

Contacts

Citation

::

@article{hvatov2023solver, AUTHOR = {Hvatov, Alexander}, TITLE = {Automated Differential Equation Solver Based on the Parametric Approximation Optimization}, JOURNAL = {Mathematics}, VOLUME = {11}, YEAR = {2023}, NUMBER = {8}, ARTICLE-NUMBER = {1787}, URL = {https://www.mdpi.com/2227-7390/11/8/1787}, ISSN = {2227-7390}, DOI = {10.3390/math11081787} }