Data-driven Thiele Equation Approach for Efficient Prediction of Spin-Torque Vortex Oscillator Dynamics

This repository contains the replication data and visualization code accompanying the paper Data-driven Thiele Equation Approach for Efficient Prediction of Spin-Torque Vortex Oscillator Dynamics. The method implements a data-driven workflow for constructing effective Thiele equations directly from micromagnetic simulations of spin-torque vortex oscillators (STVOs). The workflow consists of: generation of micromagnetic training data, extraction of effective Thiele coefficients, interpolation of the coefficient maps, prediction of vortex dynamics using the reduced-order model, visualization of both micromagnetic and reduced-model results. The data and code fully reproduce the results presented in the accompanying publication. Repository structure: config/: configuration files core/: configuration functions and constants MMS/: micromagnetic simulations training/: extraction and fitting of effective Thiele coefficients prediction/: reduced-order simulations utils/: helper functions visualization/: plotting and analysis scripts