ElemNet is a 17-layered fully connected network for the prediction of formation energy (enthalpy) from elemental compositions only. This repository contains the model weights and a Jupyter notebook for making predictions using the ElemNet model.
Input: Takes a 2D numpy array with the rows representing different compounds, and columns representing the elemental compositions with 86 elements in the set elements- ['H', 'Li', 'Be', 'B', 'C', 'N', 'O', 'F', 'Na', 'Mg', 'Al', 'Si', 'P', 'S', 'Cl', 'K', 'Ca', 'Sc', 'Ti', 'V', 'Cr', 'Mn', 'Fe', 'Co', 'Ni', 'Cu', 'Zn', 'Ga', 'Ge', 'As', 'Se', 'Br', 'Kr', 'Rb', 'Sr', 'Y', 'Zr', 'Nb', 'Mo', 'Tc', 'Ru', 'Rh', 'Pd', 'Ag', 'Cd', 'In', 'Sn', 'Sb', 'Te', 'I', 'Xe', 'Cs', 'Ba', 'La', 'Ce', 'Pr', 'Nd', 'Pm', 'Sm', 'Eu', 'Gd', 'Tb', 'Dy', 'Ho', 'Er', 'Tm', 'Yb', 'Lu', 'Hf', 'Ta', 'W', 'Re', 'Os', 'Ir', 'Pt', 'Au', 'Hg', 'Tl', 'Pb', 'Bi', 'Ac', 'Th', 'Pa', 'U', 'Np', 'Pu'], elemental compositon does not contain any element from ['He', 'Ne', 'Ar', 'Po', 'At','Rn','Fr','Ra']
Output: Returns a 1D numpy array with the predicted formation energy
Python version: 3.6
Prerequisite softwares: TensorFlow and Numpy (see requirements.txt in root directory)
-
data_preprocess.ipynb: Jupyter notebook that illustrates how to preprocess data before training an ElemNet model. It uses the training-data/oqmd_all-22Mar18.csv dataset as sample, other datasets can be similarly preprocessed. Please run this with the respective data file to create the training and test/validation sets before running the model. -
create_holdout.ipynb: Jupyter notebook that illustrates how to split preprocessed dataset to create training, validation and test sets for 10-fold cross validation, hold-out test set experiments, and the impact of training size experiment in paper [2]. -
dl_regressors.py: Code to run the ElemNet model for training. -
data_utils.py: Utility code for data loading. -
train_utils.py: Utility code for training the model. -
sample: A sample run folder that contains running configuration and the ElemNet trained using random split of training-data/oqmd_all-22Mar18.csv. The 'sample_model' can be used for transfer learning to replicate the results in paper [2]. -
make_prediction.ipynb: Jupyter notebook that illustrates how to make predictions using the trained ElemNet model from paper [1].
You can simply run the code by passing a sample config file to the dl_regressors.py as follows:
python dl_regressors.py --config_file sample/sample-run.config
The config file defines the loss_type, training_data_path, test_data_path, label, input_type [elements_tl for ElemNet] and other runtime parameters. For transfer learning used in paper [2], you need to set 'model_path' to the model checkpoint trained on the larger dataset (OQMD in our case) [e.g. "model_path":"sample/sample_model"] in the config file. The output log
from this sample run is provided in the sample/sample.log file.
- Extra files:
- sample/sample_model.data-00000-of-00001 (54 MB) - part of the model trained using sample config file above.
- tmp/model.ckpt.data-00000-of-00001 (54 MB) - part of the model used in the `make_prediction.ipynb'.
The code was developed by Dipendra Jha from the CUCIS group at the Electrical and Computer Engineering Department at Northwestern University and Logan Ward from the Computation Institute at University of Chicago.
Please cite the following works if you are using ElemNet model:
-
Dipendra Jha, Logan Ward, Arindam Paul, Wei-keng Liao, Alok Choudhary, Chris Wolverton, and Ankit Agrawal, “ElemNet: Deep Learning the Chemistry of Materials From Only Elemental Composition,” Scientific Reports, 8, Article number: 17593 (2018) [DOI:10.1038/s41598-018-35934-y] [PDF].
-
Dipendra Jha, Kamal Choudhary, Francesca Tavazza, Wei-keng Liao, Alok Choudhary, Carelyn Campbell, Ankit Agrawal, "Enhancing materials property prediction by leveraging computational and experimental data using deep transfer learning," Nature Communications, 10, Article number: 5316 (2019) [DOI: https:10.1038/s41467-019-13297-w] [PDF].
email: dipendra.jha@eecs.northwestern.edu, or ankitag@eecs.northwestern.edu
Copyright (C) 2019, Northwestern University.
See COPYRIGHT notice in top-level directory.
This work was performed under the following financial assistance award 70NANB14H012 and 70NANB19H005 from U.S. Department of Commerce, National Institute of Standards and Technology as part of the Center for Hierarchical Materials Design (CHiMaD). Partial support is also acknowledged from DOE awards DE-SC0014330, DE-SC0019358.