stefanocarrera/autophagycode_D_he_train-mercury_Qwen3-4B_strategy_trust_t1.25_g5_run2_metrics
收藏资源简介:
该数据集包含编程任务的评估数据,涵盖任务ID、入口点、可执行性、正确性、测试通过和失败数量、错误类型,以及代码复杂度度量(如Halstead词汇量、长度、体积、难度、努力值、时间,圈复杂度)、维护性指数、代码行数(LOC和SLOC)、注释百分比、TTR(类型-标记比)、标记字典、香农熵、平均和最大预测熵、定义函数数量、入口点重复等特征。数据用于分析代码质量、测试性能和编程任务表现,适用于代码分析、自动化评估和教育研究。
This dataset contains evaluation data for programming tasks, covering features such as task ID, entry point, executability, correctness, counts of passed and failed tests, error types, code complexity metrics (e.g., Halstead vocabulary size, 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, average and maximum predicted entropy, number of defined functions, duplicate entry points, and more. The data is used to analyze code quality, test performance and programming task performance, and is applicable to code analysis, automated evaluation and educational research.




