EAGER: Preserve/Destroy Decisions for Simulation Data in Computational Physics and Beyond

点击翻译按钮获取中文摘要

The scientific research community has been increasingly developing ways to share and re-use research data, thereby allowing more discoveries to be made from previous research investments. Much of the focus on sharing and reusability has been on experimental and observational data. This project addresses the equally vexing challenge of how to make best use and re-use of the massive data produced in computational simulations. Important research questions guiding this project include: the degree to which simulation results can be replicated; the advantages of storing the simulation data itself for others to reuse as compared to providing the computational software so that others can re-run the simulations; and understanding which software testing practices can facilitate the replication/reuse of simulation data and the simulation software that produces those data. The principal investigators will address these questions by performing extensive replication and software code testing on a set of computational physics simulation datasets and software code that they had gathered through a previous study. The project will produce publicly available, fully reproducible computational physics works as examples for publishing results in a way that the data and code are effectively reusable.The principal investigators aim to improve understanding of, and increase, the reusability of the code and data associated with simulation-based research. This project specifically aims to better inform data destroy/preservation decisions in the simulation context, toward improving the reusability and interoperability of simulation data and code. The project will also consider important questions such as how software engineering testing practices relate to computational physics practices, and how changes in computational environments affect code execution and the regeneration of simulation data. Ultimately, the results of this work are intended to guide the research community on how to best produce and disseminate research code. It is anticipated that the results for computational physics can be extended to develop general guidelines for simulation data and code sharing for other communities, the appropriate code testing to do so, and best practices for development of associated cyberinfrastructure and tools. This project is supported by the National Science Foundation's Public Access Initiative which is managed by the NSF Office of Advanced Cyberinfrastructure on behalf of the Foundation.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.