scikit-learn-contrib/category_encoders

Handle missing in one hot encoder

Open

#400 创建于 2023年3月12日

在 GitHub 查看
 (3 评论) (0 反应) (0 负责人)Python (2,322 star) (397 fork)batch import
buggood first issue

描述

Expected Behavior

Currently, handle_missing=value adds a new column although the documentation says 'value' will encode a new value as 0 in every dummy column. Furthermore, we need a test for this

Actual Behavior

adds a column instead of using all 0

Steps to Reproduce the Problem

from category_encoders import OneHotEncoder
import pandas as pd

he = OneHotEncoder(handle_missing="value")

data = [("foo", 1), ("bar", 2), (None, 6)]
data = pd.DataFrame(data, columns=["c1", "c2"])
print(he.fit_transform(data))

Specifications

  • Version: 2.6
  • Platform: linux

贡献者指南

技术栈
pythonpandas
领域
datatesting
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。
1-3 hours
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
Basic PythonPandasUnderstanding of one hot encodingFamiliarity with scikit learn API
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。
60
研究方向
Look at the OneHotEncoder implementation, likely in category encoders/encoders/one hot.py, to find how handle missing='value' is handled. The bug is that it adds a new column instead of filling 0 in all dummy columns. Write a test that reproduces the issue using the provided example, then fix the logic so that missing values are encoded as 0 in all columns. Also ensure that the existing tests pass. Update documentation if needed.