ART_data_analyzer: Automating parallelized computations to study the evolution of materials

The kinetics and dynamic evolution of material structures need a comprehensive understanding of the potential energy landscape at current sample state. The Activation–Relaxation Technique (ART) is an efficient way to probe the potential energy landscape by sampling a large amount of events (a single event involves initial, saddle and final state) from which a statistical distribution of activation energy barrier can be extracted. However, there has been a lack of a user-friendly toolkit to automate the parallelization of running of ART simulations and post-processing of data from ART simulations to extract useful physics information and insights. The ART_data_analyzer Python package has been developed to serve this purpose and fill in this gap for the broad community of scientific researchers interested in the kinetics and dynamic transitions of material structures. As a demo, we utilized this software package to demonstrate the user-friendly workflow of studying ZrCuAl metallic glass sample prepared by molecular dynamics. Keywords: Activation and relaxation techniques, Kinetics, Automation and parallelization, Machine learning