pymc-devs/pymc

Investigate failing test `test_parallel_sampling.py::test_abort`

Open

#4,954 创建于 2021年8月19日

在 GitHub 查看
 (0 评论) (0 反应) (0 负责人)Python (7,926 star) (1,902 fork)batch import
help wantedtests

描述

This cropped up in #4952, where we marked it as skip (https://github.com/pymc-devs/pymc3/pull/4952/commits/6249cf5082e7adb9c62ac78debb1cc20b8de5d9d) because it passed in one of the runs but failed in the next one.

It fails locally for me as well.

贡献者指南

技术栈
python
领域
machine learning
议题类型
test
难度面向新贡献者的预计实现难度，1 表示很小改动，5 表示专家级工作。
3
预计时间有经验贡献者完成调查、实现、测试并准备 pull request 的粗略时间范围。
half day
活动状态议题当前的可参与程度：新鲜、活跃、陈旧、阻塞或等待维护者输入。
stale
清晰度议题是否清楚说明期望改动、验收标准和下一步。
needs investigation
前置要求
Basic Python testing knowledgeFamiliarity with pytestUnderstanding of parallel execution
新手友好度1-100 的估计分数，表示该议题对首次贡献者的友好程度。
50
研究方向
Start by examining the test file `test parallel sampling.py`, specifically the `test abort` function. Review the related PR #4952 and the skip commit (6249cf5) for context. Run the test locally with verbose output to capture the failure. Investigate potential race conditions in the parallel sampling code, looking at files like `pymc3/sampling.py` or `pymc3/parallel sampling.py`.