Dataset and codes for: Analysis of Magnetocrystalline Anisotropy from Atom Pair Statistics in Tetragonally Strained Fe-Based Alloys: A First-Principles and Machine Learning Study

This Zenodo record contains the data and standalone plotting scripts used to reproduce paper Figures 3–8. The data are derived VASP outputs (CSV). The CSV files include spin–orbit energy differences, orbital moments (Lx/Ly/Lz), mT and exchange splitting Δ, and bond statistics from neighbor analysis; the full extraction workflow and implementation details are documented in the README. For size and reproducibility reasons, raw VASP outputs (OUTCAR/PROCAR/vasprun.xml, etc.) are not included. Contents: Data: FeSi/ and FeX_SiGeGaAl/ CSV files Utilities: data_utils.py, feature_utils.py Plot scripts: fig3_*.py–fig8_*.py, which generate the corresponding PDF figures Environment and reproduction: Python 3.8+ with numpy, pandas, matplotlib, scikit-learn environment.yml (conda) and README instructions are provided Run the scripts from this directory (e.g., python3 fig*.py) to reproduce Figures 3–8