angel

该数据集包含了来自天使系统的测试用例执行记录、源代码指标和覆盖信息，涵盖了308个构建版本，其中构建失败率为40%。数据集还包括了多种特征，如所有者经验、测试用例的年龄、所有提交者的经验、上次执行时间以及最近平均执行时间等，这些特征被用于训练学习排序（LTR）模型。该数据集收集了150个特征，并在选定的构建版本中包含了35个测试用例，其任务是测试用例优先级排序（Tcp）。

This dataset contains test case execution records, source code metrics, and coverage information from the Angel System, covering 308 build versions with a 40% build failure rate. It also includes multiple features such as owner experience, test case age, experience of all submitters, last execution time, and recent average execution time, which are utilized for training Learning to Rank (LTR) models. A total of 150 features are collected in this dataset, and 35 test cases are included in the selected build versions, with the task being Test Case Prioritization (TCP).