Aviary

The aviary contains:

- roost,
- wren,
cgcnn.

The aim is to contain multiple models for materials discovery under a common interface

Environment Setup

To use aviary you need to create an environment with the correct dependencies. The easiest way to get up and running is to use anaconda. A cudatoolkit=11.1 environment file is provided environment-gpu-cu111.yml allowing a working environment to be created with:

conda env create -f environment-gpu-cu111.yml

If you are not using cudatoolkit=11.1 or do not have access to a GPU this setup will not work for you. If so please check the following pages PyTorch, PyTorch-Scatter for how to install the core packages and then install the remaining requirements as detailed in requirements.txt.

The code was developed and tested on Linux Mint 19.1 Tessa. The code should work with other Operating Systems but it has not been tested for such use.

Aviary Setup

Once you have set up an environment with the correct dependencies you can install aviary using the following commands from the top of the directory:

conda activate aviary
python setup.py sdist
pip install -e .

This will install the library in an editable state allowing for advanced users to make changes as desired.

Example Use

To test the input files generation and cleaning/canonicalization please run:

python examples/inputs/poscar2df.py

This script will load and parse a subset of raw POSCAR files from the TAATA dataset and produce the datasets/examples/examples.csv file used for the next example. The raw files have been selected to ensure that the subset contains all the correct endpoints for the 5 elemental species in the Hf-N-Ti-Zr-Zn chemical system. All the models used share can be run on the input file produced by this example code. To test each of the three models provided please run:

python examples/roost-example.py --train --evaluate --data-path examples/inputs/examples.csv --targets E_f --tasks regression --losses L1 --robust --epoch 10

python examples/wren-example.py --train --evaluate --data-path examples/inputs/examples.csv --targets E_f --tasks regression --losses L1 --robust --epoch 10

python examples/cgcnn-example.py --train --evaluate --data-path examples/inputs/examples.csv --targets E_f --tasks regression --losses L1 --robust --epoch 10

Please note that for speed/demonstration purposes this example runs on only ~68 materials for 10 epochs - running all these examples should take < 30s. These examples do not have sufficient data or training to make accurate predictions, however, the same scripts have been used for all experiments conducted.

Cite This Work

If you use this code please cite the relevant work:

Predicting materials properties without crystal structure: Deep representation learning from stoichiometry. [Paper] [arXiv]

@article{goodall2020predicting,
  title={Predicting materials properties without crystal structure: Deep representation learning from stoichiometry},
  author={Goodall, Rhys EA and Lee, Alpha A},
  journal={Nature Communications},
  volume={11},
  number={1},
  pages={1--9},
  year={2020},
  publisher={Nature Publishing Group}
}

Rapid Discovery of Novel Materials by Coordinate-free Coarse Graining. [arXiv]

@article{goodall2021rapid,
  title={Rapid Discovery of Novel Materials by Coordinate-free Coarse Graining},
  author={Goodall, Rhys EA and Parackal, Abhijith S and Faber, Felix A and Armiento, Rickard and Lee, Alpha A},
  journal={arXiv preprint arXiv:2106.11132},
  year={2021}
}

Crystal Graph Convolutional Neural Networks for an Accurate and Interpretable Prediction of Material Properties. [Paper] [arXiv]

@article{xie2018crystal,
  title={Crystal graph convolutional neural networks for an accurate and interpretable prediction of material properties},
  author={Xie, Tian and Grossman, Jeffrey C},
  journal={Physical review letters},
  volume={120},
  number={14},
  pages={145301},
  year={2018},
  publisher={APS}
}

Disclaimer