This repo contains all codes related to the work entitled "Decoding the Pair Distribution Function of Uranium in Molten Fluoride Salts from X-ray Absorption Spectroscopy Data by Machine Learning" https://pubs.acs.org/doi/abs/10.1021/acs.jpcc.4c01898
CNN_EXAFS_PDF
| CNN sweep_WB.ipynb
script for tuning hyperparameters of our neural network using W&B
| NN_300_partial_full_multi_final.ipynb
Main notebook to execute a neural network for predicting g(r) and validating the results using experimental data
|__MD_data
| gr_output*.csv
PDFs calculated from MD structures
| output_ave_test_kx_MD.csv
FEFF-EXAFS calculated using MD structures
| r_*.txt
rmesh
|__ONNE_particle_generator
| gr.csv
| gr.png
| config.toml
configuration file for ONNE particle data construction
| rmesh.txt
| run.slurm
script to run the simulation/particle generator on supercluster
| trainingset_gen_mpi_mygr.py
main script to run ONNE particle construction
|__Scripts
Contains utility scripts for generating particles and calculating EXAFS
|__config
| config_1116023.yaml
trained parameters for the neural networks.
|__cbam/rebinE0/lightning_logs
checkpoints of neural network model
|__version_*
|__checkpoints
|__multi_tasks_FFFF
FEFF calculation using multi-processing
| config_wr.toml
configuration file for writing FEFF input files from xyz files and running FEFF calculations
| output_ave_test_kx.csv
| run.slurm
| run_FEFF.py
main script for run FEFF and write FEFF input files
| template.inp
|__toolbox
| average.py
calculate average EXAFS for all gr configurations
| average_gr.py
calculate averaged gr for all gr configurations
| check.py
Check if the calculations are complete
| run_tool.sh
| srun.sh
|__dataset_partial
gr/EXAFS training data
| gr_*.csv
| gr.png
| kmesh.txt
| rmesh.txt
| output_ave_test_kx.csv
| output_rebin.csv
Contributors:
- multi_tasks_FEFF(also see project):
- ONNE_particle_generator(also see the original repo: credits to @Nicholas Marcella Xron repo):
- neural network construction:
- @Kaifeng Zheng
- combinatorial method
- @Kaifeng Zheng