Software and the Scientist: Coding and Citation Practices in Geodynamics
Software and the Scientist: Coding and Citation Practices in Geodynamics
复制标题
软件和科学家：地球动力学中的编码和引用实践
DOI：
10.1002/2016ea000225
复制
发表时间：
2017
影响因子：
3.1
通讯作者：
L. Kellogg
中科院分区：
文献类型：
--
作者：
L. Hwang;Allison Fish;Laura Soito;MacKenzie Smith;L. Kellogg
在地球动力学和其他科学领域一样，计算已经成为研究的核心组成部分，补充了实地观察，实验室分析，实验和理论。用于数据分析、绘图、可视化、建模和模拟的计算工具对于科学工作流程的各个方面都是必不可少的。专门的科学软件通常是由地球动力学家开发供自己使用的，这种努力代表了一种独特的智力贡献。借鉴地球动力学社区，专注于开发和传播科学软件，我们评估软件开发和属性的当前实践，以及对软件引用的需要和最佳实践的态度。我们分析了地球动力学计算基础设施参与者的出版物，并对固体地球物理学界进行了混合方法调查。由此我们了解到，编码技能通常是非正式学习的。参与者认为好的代码是可信的，可重用的，可读的，并且不太复杂，并认为一个好的编码器是一个以开放和合理的庄园参与社区的人，为长期和短期社区项目做出贡献。参与者强烈支持引用软件，这反映在文献中命名软件包的高比率和参考文献中的高引用率。然而，缺乏开发人员关于如何引用的明确指示和用户关于引用内容的教育。此外，引用并不总是导致资源的可重复性。软件包本身的唯一标识符、社区教育和引用工具将有助于更好的归因实践。
In geodynamics as in other scientific areas, computation has become a core component of research, complementing field observation, laboratory analysis, experiment, and theory. Computational tools for data analysis, mapping, visualization, modeling, and simulation are essential for all aspects of the scientific workflow. Specialized scientific software is often developed by geodynamicists for their own use, and this effort represents a distinctive intellectual contribution. Drawing on a geodynamics community that focuses on developing and disseminating scientific software, we assess the current practices of software development and attribution, as well as attitudes about the need and best practices for software citation. We analyzed publications by participants in the Computational Infrastructure for Geodynamics and conducted mixed method surveys of the solid earth geophysics community. From this we learned that coding skills are typically learned informally. Participants considered good code as trusted, reusable, readable, and not overly complex and considered a good coder as one that participates in the community in an open and reasonable manor contributing to both long‐ and short‐term community projects. Participants strongly supported citing software reflected by the high rate a software package was named in the literature and the high rate of citations in the references. However, lacking are clear instructions from developers on how to cite and education of users on what to cite. In addition, citations did not always lead to discoverability of the resource. A unique identifier to the software package itself, community education, and citation tools would contribute to better attribution practices.
影响因子：
3.8
作者：
Smith, Arfon M.;Katz, Daniel S.;Niemeyer, Kyle E.
通讯作者：
Niemeyer, Kyle E.