stefanocarrera/autophagycode_D_he_train-mercury_Qwen3-4B_strategy_trust_t1.25_g9_run1_metrics
收藏资源简介:
该数据集是一个包含代码任务相关信息的集合,主要用于分析软件质量、测试性能和代码复杂度。数据集包含164个训练示例,每个示例具有多个特征,如任务ID、入口点、是否可执行、是否正确、测试通过和失败数量、测试运行时间(当前为空)、错误类型。此外,还包括Halstead软件度量指标(如词汇量、长度、体积、难度、努力、时间)、圈复杂度、可维护性指数、代码行数(LOC和SLOC)、注释百分比、TTR(类型-令牌比率)、令牌字典、香农熵、平均和最大预测熵、定义函数数量以及入口点是否重复。这些特征有助于评估代码的可读性、维护性和测试效果,适用于软件工程、机器学习和代码分析研究。
This dataset is a collection of information related to code tasks, primarily used for analyzing software quality, testing performance, and code complexity. It contains 164 training examples, each with multiple features such as task ID, entry point, executability, correctness, number of tests passed and failed, test run time (currently null), and error type. Additionally, it includes Halstead software metrics (e.g., vocabulary, length, volume, difficulty, effort, time), cyclomatic complexity, maintainability index, lines of code (LOC and SLOC), comment percentage, TTR (Type-Token Ratio), token dictionary, Shannon entropy, mean and max predictive entropy, number of functions defined, and whether the entry point is repeated. These features are useful for evaluating code readability, maintainability, and testing effectiveness, making the dataset suitable for research in software engineering, machine learning, and code analysis.




