stefanocarrera/autophagycode_D_he_train-mercury_Qwen3-8B_strategy_trust_t1.25_g2_run0_metrics
收藏资源简介:
该数据集是一个用于代码质量评估和编程任务分析的数据集,包含164个训练样本。数据集特征包括任务ID(task_id)、入口点(entry_point)、可执行性(is_executable)、正确性(is_correct)、测试通过和失败数量(tests_passed, tests_failed),以及代码度量指标如Halstead复杂度(halstead_vocabulary, halstead_length, halstead_volume, halstead_difficulty, halstead_effort, halstead_time)、圈复杂度(cyclomatic_complexity)、可维护性指数(maintainability_index)、代码行数(loc, sloc)、注释百分比(comment_percentage)、TTR(token type ratio)、香农熵(shannon_entropy)、预测熵(mean_predictive_entropy, max_predictive_entropy)、函数定义数量(n_func_defined)、入口点重复性(entry_point_repeated)等。数据用于分析编程任务的执行结果和代码复杂度。
This dataset is designed for code quality assessment and programming task analysis, containing 164 training examples. It includes features such as task ID, entry point, executability (is_executable), correctness (is_correct), number of tests passed and failed (tests_passed, tests_failed), and code metrics like Halstead complexity measures (halstead_vocabulary, halstead_length, halstead_volume, halstead_difficulty, halstead_effort, halstead_time), cyclomatic complexity, maintainability index, lines of code (loc, sloc), comment percentage, TTR (token type ratio), Shannon entropy, predictive entropy (mean_predictive_entropy, max_predictive_entropy), number of functions defined (n_func_defined), and entry point repetition (entry_point_repeated). The data is used to analyze execution outcomes and code complexity in programming tasks.




