combined-gnn

KO | EN

gitlite — search

#javascript #python #hacktoberfest #react #ai #typescript #llm #go #golang #android #machine-learning #rust #deep-learning #linux

★ 10 Open GitHub ↗

Graph neural network for predicting energy of known and hypothetical crystal structures

// repository documentation

Was this content helpful?

★ 0 (0 ratings)

## GNN For Predicting Energy of Known and Hypothetical Structures

This package provides the following functionalities: 
* Predict total energy of new crystal strtuctures using pre-trained models
* Train models with custom datasets

### Package Requirements

Following python packages are needed:

* [TensorFlow 2](https://www.tensorflow.org/install) 
* [TensorFlow Addons](https://www.tensorflow.org/addons/overview) 
* [pymatgen](https://pymatgen.org/installation.html) 
* [scikit-learn](https://scikit-learn.org/stable/install.html) 
* [nfp](https://pypi.org/project/nfp/)

### Usage

* Predict total energy using pre-trained models: `energy_prediction_demo` 
* Train a model using a custom dataset: `train_model`
* Pre-trained models: `pretrained_models`

### Datasets

To reproduce the models, download the following datasets:

* *Total Energy*

* [Total energy of ICSD structures from NRELMatDB database (materials.nrel.gov)](nrelmatdb_icsd_energies.csv) 
  * [Total energy of hypothetical structures](hypothetical_structure_energies.csv)

* *Crystal Structures*  
  
  * ICSD crystal structure files cannot be distributed 
  * [Relaxed structures of hypothetical materials](relaxed_hypothetical_structures.tar.gz)

### Authors

The files in this repository are curated and maintained by

* [Shubham Pandey](mailto:shubhampandey[at]mines[dot]edu)
* [Peter St. John](mailto:Peter.STJohn[at]nrel[dot]gov)
* [Prashun Gorai](mailto:pgorai[at]mines[dot]edu)

### Cite
```
@article{pandey2021,
  title={Predicting energy and stability of known and hypothetical crystals using graph neural network},
  author={Pandey, Shubham and Qu, Jiaxing and Stevanovi{\'c}, Vladan and John, Peter St and Gorai, Prashun},
  journal={Patterns},
  volume={2},
  pages={100361},
  year={2021},
  publisher={Elsevier},
  url = {https://doi.org/10.1016/j.patter.2021.100361},
  doi = {10.1016/j.patter.2021.100361}
}
```

### License

This package is released under the MIT License.

combined-gnn

Explore Similar Repositories

Was this content helpful?

Recent Feedback

Download README