stefanocarrera/autophagycode_D_he_train-mercury_Qwen3-8B_strategy_trust_t1.25_g1_run1_metrics
收藏资源简介:
这是一个代码执行和测试评估数据集,包含编程任务的相关信息。数据集记录了每个任务的ID、入口点、可执行性、正确性、通过和失败的测试数量,以及测试运行时间(当前为空)。此外,它提供了代码复杂度度量,如Halstead指标(词汇量、长度、体积、难度、努力程度、时间)、圈复杂度、可维护性指数,以及代码行数(LOC和SLOC)、注释百分比、TTR(类型-标记比率)、令牌字典、香农熵、平均和最大预测熵、定义函数数量和入口点重复情况。数据集用于分析和评估代码质量、测试性能和复杂度,适用于代码分析、机器学习和软件工程研究。
This is a code execution and testing evaluation dataset containing information related to programming tasks. It records each tasks ID, entry point, executability, correctness, number of tests passed and failed, and test run time (currently null). Additionally, it provides code complexity metrics such as Halstead measures (vocabulary, length, volume, difficulty, effort, time), cyclomatic complexity, maintainability index, as well as code lines (LOC and SLOC), comment percentage, TTR (Type-Token Ratio), token dictionary, Shannon entropy, mean and max predictive entropy, number of defined functions, and entry point repetition. The dataset is used for analyzing and evaluating code quality, testing performance, and complexity, suitable for code analysis, machine learning, and software engineering research.




