FastVPINNs

FastVPINNs: An efficient tensor-based Python library for solving partial differential equations using hp-Variational Physics Informed Neural Networks - Published in JOSS (2024)

https://github.com/cmgcds/fastvpinns

Science Score: 100.0%

This score indicates how likely this project is to be science-related based on various indicators:

✓
CITATION.cff file
Found CITATION.cff file
✓
codemeta.json file
Found codemeta.json file
✓
.zenodo.json file
Found .zenodo.json file
✓
DOI references
Found 1 DOI reference(s) in JOSS metadata
✓
Academic publication links
Links to: arxiv.org, joss.theoj.org
✓
Committers with academic emails
3 of 4 committers (75.0%) from academic institutions
✓
Institutional organization owner
Organization cmgcds has institutional domain (cmg.cds.iisc.ac.in)
✓
JOSS paper metadata
Published in Journal of Open Source Software

Keywords

deep-learning inverse-problems neural-network partial-differential-equations physics-informed-neural-networks pinn scientific-machine-learning tensorflow2

Scientific Fields

Engineering Computer Science - 40% confidence

Last synced: 4 months ago · JSON representation ·

Repository

FastVPINNs - A tensor-driven acceleration of VPINNs for complex geometries

Basic Info

Host: GitHub
Owner: cmgcds
License: mit
Language: Python
Default Branch: main
Homepage: https://cmgcds.github.io/fastvpinns/
Size: 38.6 MB

Statistics

Stars: 43
Watchers: 3
Forks: 67
Open Issues: 1
Releases: 3

Topics

deep-learning inverse-problems neural-network partial-differential-equations physics-informed-neural-networks pinn scientific-machine-learning tensorflow2

Created over 1 year ago · Last pushed 11 months ago

Metadata Files

Readme Changelog Contributing License Code of conduct Citation

Tensor-driven accelerated framework for hp-variational pinns

Link to Documentation 📚

A robust tensor-based deep learning framework for solving partial differential equations using hp-Variational Physics-Informed Neural Networks (hp-VPINNs). The framework is based on the methodology presented in the FastVPINNs Paper.

This library is a highly optimised version of the the initial implementation of hp-VPINNs by Kharazmi et al.. Refer the hp-VPINNs Paper.

Authors 👨‍💻

Thivin Anandh, Divij Ghose, Sashikumaar Ganesan

STARS Lab, Department of Computational and Data Sciences, Indian Institute of Science, Bangalore, India

Installation 🛠️

The build of the code is currently tested on Python versions (3.8, 3.9, 3.10, 3.11), on OS Ubuntu 20.04 and Ubuntu 22.04, MacOS-latest and Windows-latest (refer compatibility build Compatability check).

You can install the package using pip as follows:

bash pip install fastvpinns

On ubuntu systems with libGL issues caused due to matplotlib or gmsh, please run the following command to install the required dependencies: bash sudo apt-get install -y libglu1-mesa

For more information on the installation process, please refer to our documentation here.

Citing 📜

If you use this code in your research, please consider citing the following paper:

bibtex @misc{anandh2024fastvpinns, title={FastVPINNs: Tensor-Driven Acceleration of VPINNs for Complex Geometries}, author={Thivin Anandh, Divij Ghose, Himanshu Jain and Sashikumaar Ganesan}, year={2024}, eprint={2404.12063}, archivePrefix={arXiv}, primaryClass={cs.LG} }

Usage 🚀

For detailed usage, please refer to our documentation here.

The package provides a simple API to train and solve PDE using VPINNs. The following code snippet demonstrates how to train a hp-VPINN model for the 2D Poisson equation for a structured grid. We could observe that we can solve a PDE using fastvpinns using 15 lines of code.

```python

load the geometry

domain = Geometry2D("quadrilateral", "internal", 100, 100, "./") cells, boundarypoints = domain.generatequadmeshinternal(xlimits=[0, 1],ylimits=[0, 1],ncellsx=4, ncellsy=4, numboundary_points=400)

load the FEspace

fespace = Fespace2D(domain.mesh,cells,boundarypoints,domain.meshtype,feorder=5,fetype="jacobi",quadorder=5,quadtype="legendre", fetransformationtype="bilinear",boundfunctiondict=boundfunctiondict,boundconditiondict=boundconditiondict, forcingfunction=rhs,outputpath=ioutputpath,generatemeshplot=True)

Instantiate Data handler

datahandler = DataHandler2D(fespace, domain, dtype=tf.float32)

Instantiate the model with the loss function for the model

model = DenseModel(layerdims=[2, 30, 30, 30, 1],learningratedict=0.01,paramsdict=paramsdict, lossfunction=pdelosspoisson, ## Loss function of poisson2D inputtensorslist=[intensor, dirin, dirout], origfactormatrices=[datahandler.shapevalmatlist,datahandler.gradxmatlist, datahandler.gradymatlist], forcefunctionlist=datahandler.forcingfunctionlist, tensordtype=tf.float32, useattention=iuseattention, ## Archived (not in use) activation=i_activation, hessian=False)

Train the model

for epoch in range(1000): model.train_step() ```

Note : Supporting functions which define the actual solution and boundary conditions have to be passed to the main code.

Contributing 🤝

This code is currently maintained by the authors as mentioned in the section above. We welcome contributions from the community. Please refer to the documentation for guidelines on contributing to the project.

License 📑

This project is licensed under the MIT License - see the LICENSE file for details.

Owner

Name: Computational Mathematics Group (CMG)
Login: cmgcds
Kind: organization
Location: Indian Institute of Science, Bangalore, India

Website: http://cmg.cds.iisc.ac.in/
Repositories: 1
Profile: https://github.com/cmgcds

CMG at the Department of Computational and Data Sciences, IISc

JOSS Publication

FastVPINNs: An efficient tensor-based Python library for solving partial differential equations using hp-Variational Physics Informed Neural Networks

Published

July 30, 2024

DOI

10.21105/joss.06764

Volume 9, Issue 99, Page 6764

Authors

Thivin Anandh

Department of Computational and Data Sciences, Indian Institute of Science, Bangalore, India

Divij Ghose

Department of Computational and Data Sciences, Indian Institute of Science, Bangalore, India

Sashikumaar Ganesan

Department of Computational and Data Sciences, Indian Institute of Science, Bangalore, India

Editor

Patrick Diehl

Citation (CITATION.cff)

cff-version: "1.2.0"
authors:
- family-names: Anandh
  given-names: Thivin
  orcid: "https://orcid.org/0000-0003-4969-3242"
- family-names: Ghose
  given-names: Divij
  orcid: "https://orcid.org/0009-0005-6295-543X"
- family-names: Ganesan
  given-names: Sashikumaar
  orcid: "https://orcid.org/0000-0003-1858-3972"
contact:
- family-names: Anandh
  given-names: Thivin
  orcid: "https://orcid.org/0000-0003-4969-3242"
- family-names: Ganesan
  given-names: Sashikumaar
  orcid: "https://orcid.org/0000-0003-1858-3972"
doi: 10.5281/zenodo.13127234
message: If you use this software, please cite our article in the
  Journal of Open Source Software.
preferred-citation:
  authors:
  - family-names: Anandh
    given-names: Thivin
    orcid: "https://orcid.org/0000-0003-4969-3242"
  - family-names: Ghose
    given-names: Divij
    orcid: "https://orcid.org/0009-0005-6295-543X"
  - family-names: Ganesan
    given-names: Sashikumaar
    orcid: "https://orcid.org/0000-0003-1858-3972"
  date-published: 2024-07-30
  doi: 10.21105/joss.06764
  issn: 2475-9066
  issue: 99
  journal: Journal of Open Source Software
  publisher:
    name: Open Journals
  start: 6764
  title: "FastVPINNs: An efficient tensor-based Python library for
    solving partial differential equations using hp-Variational Physics
    Informed Neural Networks"
  type: article
  url: "https://joss.theoj.org/papers/10.21105/joss.06764"
  volume: 9
title: "FastVPINNs: An efficient tensor-based Python library for solving
  partial differential equations using hp-Variational Physics Informed
  Neural Networks"

GitHub Events

Total

Watch event: 23
Delete event: 4
Issue comment event: 1
Push event: 20
Pull request event: 2
Fork event: 55
Create event: 3

Last Year

Watch event: 23
Delete event: 4
Issue comment event: 1
Push event: 20
Pull request event: 2
Fork event: 55
Create event: 3

Committers

Last synced: 5 months ago

All Time

Total Commits: 152
Total Committers: 4
Avg Commits per committer: 38.0
Development Distribution Score (DDS): 0.171

Past Year

Commits: 12
Committers: 2
Avg Commits per committer: 6.0
Development Distribution Score (DDS): 0.333

Top Committers

Name	Email	Commits
thivin	t**h@i**n	126
divijghose	d**e@i**n	24
Ziyi Yin	z**n@g**u	1
Prof. Sashikumaar Ganesan	s****n	1

Committer Domains (Top 20 + Academic)

iisc.ac.in: 2 gatech.edu: 1

Issues and Pull Requests

Last synced: 4 months ago

All Time

Total issues: 4
Total pull requests: 20
Average time to close issues: 14 days
Average time to close pull requests: about 15 hours
Total issue authors: 4
Total pull request authors: 5
Average comments per issue: 3.25
Average comments per pull request: 0.2
Merged pull requests: 16
Bot issues: 0
Bot pull requests: 0

Past Year

Issues: 2
Pull requests: 8
Average time to close issues: 1 day
Average time to close pull requests: 1 day
Issue authors: 2
Pull request authors: 5
Average comments per issue: 1.5
Average comments per pull request: 0.25
Merged pull requests: 6
Bot issues: 0
Bot pull requests: 0

View more stats

Top Authors

Issue Authors

ziyiyin97 (1)
pescap (1)
tsarikahin (1)
margecruz (1)
divijghose (1)

Pull Request Authors

thivinanandh (14)
divijghose (8)
jaihimanshu (2)
ziyiyin97 (1)
danielskatz (1)

Top Labels

Issue Labels

enhancement (3)

Pull Request Labels

enhancement (1)

Packages

Total packages: 1
Total downloads:
- pypi 22 last-month

Total dependent packages: 0
Total dependent repositories: 0
Total versions: 3
Total maintainers: 1

pypi.org: fastvpinns

A fast tensor-driven variational physics-informed neural network library for solving PDEs.

Homepage: https://github.com/cmgcds/fastvpinns
Documentation: https://cmgcds.github.io/fastvpinns/
License: MIT License
Latest release: 1.0.2
published over 1 year ago

Versions: 3
Dependent Packages: 0
Dependent Repositories: 0
Downloads: 22 Last month

Rankings

Dependent packages count: 9.4%

Average: 35.7%

Dependent repos count: 62.0%

Maintainers (1)

thivinanandh

Last synced: 4 months ago

Dependencies

.github/workflows/code-quality-control.yml actions

actions/checkout v4 composite
actions/setup-python v5 composite

.github/workflows/integration-tests.yml actions

actions/checkout v4 composite
actions/setup-python v5 composite

.github/workflows/unit-tests.yml actions

actions/checkout v4 composite
actions/setup-python v5 composite

fastvpinns/FastVPINNs.egg-info/requires.txt pypi

gmsh *
matplotlib *
numba *
numpy *
pandas *
pyDOE *
pygmsh *
pytest *
pyvista *
pyyaml *
scipy *
seaborn *
sympy *
tensorflow *
tqdm *
vtk *
wandb *

pyproject.toml pypi

requirements.txt pypi

black *
gmsh *
matplotlib *
numpy ==1.21.4
pandas *
pre-commit *
pyDOE *
pygmsh *
pytest *
pyyaml *
rich ==12.5.1
scipy *
seaborn *
tensorflow ==2.9.1
tqdm *
vtk *

setup.py pypi

.github/workflows/code-coverage.yml actions

actions/checkout v4 composite
actions/setup-python v2 composite
codecov/codecov-action v4.2.0 composite

.github/workflows/compatibility-tests.yml actions

actions/checkout v3 composite
actions/setup-python v3 composite

.github/workflows/documentation-deployment.yml actions

actions/checkout v4 composite
actions/setup-python v5 composite
peaceiris/actions-gh-pages v3 composite

.github/workflows/pypi-deployment.yml actions

actions/checkout v3 composite
actions/setup-python v3 composite

FastVPINNs

Science Score: 100.0%

Keywords

Scientific Fields

Repository

Basic Info

Statistics

Topics

Metadata Files

README.md

Tensor-driven accelerated framework for hp-variational pinns

Authors 👨‍💻

Installation 🛠️

Citing 📜

Usage 🚀

load the geometry

load the FEspace

Instantiate Data handler

Instantiate the model with the loss function for the model

Train the model

Contributing 🤝

License 📑

Owner

JOSS Publication

FastVPINNs: An efficient tensor-based Python library for solving partial differential equations using hp-Variational Physics Informed Neural Networks

Authors

Editor

Tags

Citation (CITATION.cff)

GitHub Events

Total

Last Year

Committers

All Time

Past Year

Top Committers

Committer Domains (Top 20 + Academic)

Issues and Pull Requests

All Time

Past Year

Top Authors

Issue Authors

Pull Request Authors

Top Labels

Issue Labels

Pull Request Labels

Packages

pypi.org: fastvpinns

Rankings

Maintainers (1)

Dependencies