NicolasHug/Surprise

Potential numerical instability in pearson sim

Open

#216 创建于 2018年10月15日

在 GitHub 查看
 (2 评论) (0 反应) (0 负责人)Python (6,098 star) (1,043 fork)batch import
help wanted

描述

Following #214, there might be some numerical instability in pearson similarity computation, probably caused by the sqrt function receiving either negative values, NaN or infinity.

I couldn't reproduce the issue myself.

贡献者指南

技术栈
python
领域
machine learning
议题类型
bug
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。
1-2 days
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
Pythonbasic linear algebranumpyfloating point arithmetic
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。
25
研究方向
Investigate the Pearson similarity computation in the Surprise library, likely in the file `surprise/similarities.py` or similar. Look for the sqrt function call and ensure it doesn't receive negative values or NaN. Check the linked issue #214 for context. Attempt to reproduce the instability by testing with datasets that have zero variance or near zero similarities. Consider adding a small epsilon or clamping values before sqrt.