基于复杂网络的开源软件生态系统研究——以R软件为例

A Research on the Open Source Ecosystem Based on Complex Networks:A Case Study for R Language

R软件作为统计领域重要的开源软件,开发历史久远,生态系统较为成熟.基于此,文章从复杂的有向网络角度出发,探讨R软件包的核心组成、依赖关系,对软件包之间的依赖网络进行社区划分,对其发展脉络进行深入研究,从而可以让R语言开发者、使用者快速了解R软件.研究表明:R软件迅速发展,功能多样;R软件包间的依赖关系服从幂律分布和“小世界”现象;R软件包的依赖网络中有“统计模型”,“高性能计算”,“数据可视化”,“网页技术”,“数据预处理”,“生物信息”子社区.综合上述分析,文章有如下发现:R软件生态系统可满足整个数据分析全流程的需求,与时俱进地扩展,吸引了来自世界各地的开发者,注重长期维护软件包的健康.最后,文章从R软件的开源模式中来探索对国产开源软件及其生态系统的指导意义.

As an important open source software in the field of statistics,R language has a long history of development and a relatively mature ecosystem.This paper explores the core components and dependencies of R language packages from the perspective of complex directed networks,divides the dependency networks between packages,and conducts an in-depth study of their developmental lineage,so that R language developers and users can quickly understand R language.Our study shows that the dependency relationships among the R packages obey the power-law distribution and“small-world”phenomenon;the dependency network of R packages includes five subcommunities“Statistical Modeling”,“High-Performance Computing”,“Data Visualization”,“Statistical Modeling”,“Data Visualization”,“Web technologies”,“Bioinformatics”.In summary,this paper has found the following:The R language ecosystem meets the needs of the entire data analysis process,scales with the times,attracts developers from all over the world,and focuses on the long-term maintenance of the package's health.Finally,the paper explores the open source model of R language to explore the implications for domestic open source software and its ecosystem.