Floating-Point Functions Dataset
收藏资源简介:
本数据集为首个大规模实证研究,专注于静态类型语言中浮点运算的使用情况,数据来源于公共GitHub仓库。研究采用了包括随机抽样和基于代码内在属性过滤在内的先进挖掘技术,以避免偏差。数据集通过在源代码中搜索关键词和分析代码结构来识别浮点运算的使用。该数据集为10百万个真实世界的浮点函数,为构建新基准和评估浮点分析技术奠定了基础,具有高度的代表性,能够为浮点运算自动化推理技术的开发与评估提供支持。
This dataset is the first large-scale empirical study focusing on the usage of floating-point operations in statically typed languages. The data is sourced from public GitHub repositories. The study adopts advanced mining techniques including random sampling and filtering based on inherent code properties to mitigate bias. It identifies the usage of floating-point operations by searching for keywords in source code and analyzing code structures. Comprising 10 million real-world floating-point functions, this dataset lays a solid foundation for establishing new benchmarks and evaluating floating-point analysis technologies, featuring high representativeness that supports both the development and assessment of automated reasoning techniques for floating-point operations.




