基于一列作为变量，两列作为值，将 pandas df 从长转换为宽

Question

我有一个数据框

{'Author': {0: 1, 1: 1, 2: 2, 3: 2},
'Article': {0: 11, 1: 11, 2: 22, 3: 22},
'Year': {0: 2017, 1: 2018, 2: 2017, 3: 2018},
'First': {0: 1, 1: 0, 2: 0, 3: 0},
'Second': {0: 0, 1: 1, 2: 1, 3: 1}}

想要为 Year 从长转换为宽，创建基于 First 和 Second 的值列。

预期输出

Author  Article Year    First   Second  First_2017  First_2018  Second_2017 Second_2018
1        11     2017      1      0         1          0             0          1
1        12     2018      0      1         1          0             0          1
2        22     2017      0      0         0          0             0          1
2        23      2018     0      1         0          0             0          1

Answer 1

如果需要测试是否在 ['First','Second'] 列中至少存在一个 1，请使用 DataFrame.pivot_table 和 any，展平 MultiIndex 并附加到原始：

df1 = df.pivot_table(index='Author', 
                     columns='Year', 
                     values=['First','Second'], 
                     aggfunc='any')
df1.columns = [f'{a}_{b}' for a, b in df1.columns]

df = df.join(df1.astype(int), on='Author')
print (df)
   Author  Article  Year  First  Second  First_2017  First_2018  Second_2017  \
0       1       11  2017      1       0           1           0            0   
1       1       11  2018      0       1           1           0            0   
2       2       22  2017      0       1           0           0            1   
3       2       22  2018      0       1           0           0            1   

   Second_2018  
0            1  
1            1  
2            1  
3            1

Answer 2

IUUC，你可以pivot and merge:

df2 = df.pivot(index=['Author', 'Article'], columns='Year')
df2.columns = df2.columns.map(lambda x: '_'.join(map(str, x)))
df.merge(df2, left_on=['Author', 'Article'], right_index=True)

输出：

   Author  Article  Year  First  Second  First_2017  First_2018  Second_2017  Second_2018
0       1       11  2017      1       0           1           0            0            1
1       1       11  2018      0       1           1           0            0            1
2       2       22  2017      0       1           0           0            1            1
3       2       22  2018      0       1           0           0            1            1

基于一列作为变量，两列作为值，将 pandas df 从长转换为宽

Convert pandas df from long to wide based on one column as variable and two columns for values

python

pandas

melt