scikit-learn-contrib/category_encoders

Metadata for encoders

Open

#166 创建于 2019年1月5日

在 GitHub 查看
 (10 评论) (2 反应) (0 负责人)Python (2,322 star) (397 fork)batch import
enhancementhelp wanted

描述

It should be possible to programmatically differentiate between encoders that:

  1. Do not require any target during fitting (like OneHotEncoder).
  2. Require some target during fitting (like TargetEncoder).
  3. Require binary target during fitting (like WOE).

This information would be useful for:

  1. Parameterized tests.
  2. Users that wonder, which encoder they may use.

Proposed implementation: Like in scikit-learn.

贡献者指南

技术栈
python
领域
machine learning
议题类型
feature
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。
clear
前置要求
Python programmingscikit learn API knowledgecategory encoders library structure
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。
70
研究方向
Examine the existing encoder classes in category encoders to identify their target requirements. Look at scikit learn's Estimator type tags for inspiration on how to implement a metadata system. Review the issue comments for additional design considerations and potential pitfalls. The implementation could involve adding a class attribute or method to each encoder indicating target dependency.