具有重复项的类的多标签编码 [英] Multi label encoding for classes with duplicates

问题描述

如何对包含重复项的列表列进行n-hot编码?

How can I n-hot encode a column of lists with duplicates?

类似于sklearn的MultiLabelBinarizer，它计算重复类的实例数量而不是二值化.

Something like MultiLabelBinarizer from sklearn which counts the number of instances of duplicate classes instead of binarizing.

示例输入:

x = pd.Series([['a', 'b', 'a'], ['b', 'c'], ['c','c']])

预期输出:

    a   b   c
0   2   1   0
1   0   1   1
2   0   0   2

推荐答案

我基于

import pandas as pd
x = pd.Series([['a', 'b', 'a'], ['b', 'c'], ['c','c']])

mlc = MultiLabelCounter()
mlc.fit_transform(x)

# [[2, 1, 0], [0, 1, 1], [0, 0, 2]]

这篇关于具有重复项的类的多标签编码的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！