Fundamental Microscopic Properties as Predictors of Large-Scale Quantities of Interest: Validation through Grain Boundary Energy Trends

Supporting data and code for (https://arxiv.org/abs/2411.16770)

Authors: Benjamin A. Jasperson, Ilia Nikiforov, Amit Samanta, Brandon Runnels, Harley T. Johnson, Ellad B. Tadmor

Installation

1. Create a conda environment using the attached yml file.
2. Activate conda environment.
3. Install wield per the instructions. This research utilized the 51f96b4 commit, which should install in the conda environment provided.

Descriptions

Dataset and script descriptions.

Datasets

• df_analytical.csv: lattice-matching model GB curves.
• df_md_avg.csv: averages of IP-generated GB curves and property data from OpenKIM.
• df_merge_all.csv: combined dataset of OpenKIM IP and analytical results, with outlier removal and filtering applied (see data_import.py for workflow). Includes scaling coefficient.
• df_merge_raw.csv: raw dataset from OpenKIM, prior to data_import workflow.
• df_merge.csv: subset of df_merge_all, filtered by tilt axis and duplicates removed.
• dft.csv: dft dataset.
• gb_dft_combined.csv: merge individual lines from gb_dft.csv so they are grouped by species and tilt axis.
• gb_dft.csv: original GB DFT results, from Crystalium database.
• prop_table.csv: property descriptions for manuscript.
• stats.csv: database stats.

Scripts

• bibfile_create.py: used to generate bibfile/citations for all IPs used.
• data_exploration.py: plotting and analysis for manuscript/supplemental. Includes GB plots for manuscript/supplemental (Fig 1), correlation heatmap w/ barchart (Fig 2), pairplots (Figs 3 and 7).
• data_import.py: combine OpenKIM property data, analytical results, and grain boundary data. Save combined results for further analysis.
• dft_import.py: import and analysis of DFT results and predictions (Fig 6)
• generate_analytical.py: produce GB lattice matching interatomic energy model
• model_selection.py: model evaluation using k-fold CV. Factor importance figure for manuscript (Fig 4).
• models.py: various helper functions
• openKimInterface.py: helper code to extract property data from OpenKIM database.
• plotting.py: plotting code
• uncertainty_quantification_nested_cv.py: k-fold cross-validation, linear and SVR models (Fig 5)