stefanocarrera/autophagycode_D_he_train-mercury_Qwen3-4B_strategy_trust_t1.1_g10_run0_metrics
收藏资源简介:
该数据集包含关于代码任务执行与度量的结构化信息,主要用于分析和评估代码质量与复杂性。数据集的字段(features)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/失败数量(tests_passed/tests_failed)、错误类型(error_type),以及多种代码度量指标,如Halstead复杂度(包括词汇量、长度、体积、难度、努力程度和时间)、圈复杂度(cyclomatic_complexity)、可维护性指数(maintainability_index)、代码行数(loc和sloc)、注释比例(comment_percentage)、类型-标记比(TTR)、标记字典(token_dict)、香农熵(shannon_entropy)、预测熵(mean_predictive_entropy和max_predictive_entropy)、定义函数数量(n_func_defined)和入口点重复性(entry_point_repeated)。数据集仅包含一个训练集(train),共有164个样本,总大小约为232KB,适用于代码分析、机器学习模型训练或软件工程研究等任务。
This dataset contains structured information on code task execution and metrics, primarily for analyzing and evaluating code quality and complexity. The datasets features include task ID (task_id), entry point (entry_point), executability (is_executable), correctness (is_correct), number of tests passed/failed (tests_passed/tests_failed), error type (error_type), and various code metrics such as Halstead complexity (including vocabulary, length, volume, difficulty, effort, and time), cyclomatic complexity (cyclomatic_complexity), maintainability index (maintainability_index), lines of code (loc and sloc), comment percentage (comment_percentage), type-token ratio (TTR), token dictionary (token_dict), Shannon entropy (shannon_entropy), predictive entropy (mean_predictive_entropy and max_predictive_entropy), number of functions defined (n_func_defined), and entry point repetition (entry_point_repeated). The dataset includes only a training set (train) with 164 examples and a total size of approximately 232KB, suitable for tasks like code analysis, machine learning model training, or software engineering research.




