Pandas多值分类变量如何进行二进制编码？

Question

我有以下数据帧，其中某列有多个值：

             my column

         0 - ["A", "B"]
         1 - ["B", "C", "D"]
         2 - ["B", "D"]

如何获得这样的数据帧：（其中每个列都采用“我的列”中的值的名称）

         "A"  "B"  "C"  "D"
      0 - 1    1    0    0
      1 - 0    1    1    1
      2 - 0    1    0    1

Answer 1

为了好玩，这里有一个傻瓜的天真实现：

import pandas as pd

my_column = pd.Series([['A','B'],['B','C','D'],['B','D']])

frameA = pd.DataFrame(my_column, columns=['my_column'])
#extract all new headers from the DataFrame rows, in order:
headers = sorted(list(set([x for y in frame['my_column'] for x in y])))
#make a list of the DataFrame rows (stored as lists):
rows = [y for x in range(len(frame)) for y in frame.loc[x]]

builder = {}               #construct a dictionary to build a new DataFrame from
for header in headers:
    column = []
    for row in rows:
        if header in row:
            column.append(1)
        else:
            column.append(0)
    builder.update({header:column})

frameB = pd.DataFrame(builder)

print(frameB)

结果：

   A  B  C  D
0  1  1  0  0 
1  0  1  1  1
2  0  1  0  1