使用 Pivot/melt 转换 pandas 数据

Question

我有像

import pandas as pd 

pd.DataFrame({'Date': ["2022-01-01","2022-01-02","2022-01-03"],
         'customer_type': ['A','A','B'],
         'process_A': [2,4,5],
         'process_B': [3,9,6]})

op
      Date         customer_type    process_A   process_B
0     2022-01-01    A               2           3
1     2022-01-02    A               4           9
2     2022-01-03    B               5           6

我希望它以这样的方式转换的数据框，即日期变成列，数据框看起来像这样

  customer_type    Process  2022-01-01  2022-01-02  2022-01-03
0             A  process_A           2           4           0
1             A  process_B           3           9           0
2             B  process_A           0           0           5
3             B  process_B           0           0           6

Answer 1

您可以组合熔体和pivot_table：

(df
 .melt(id_vars=['Date', 'customer_type'], var_name='Process')
 .pivot_table(index=['customer_type', 'Process'], columns='Date',
              values='value', fill_value=0)
 .reset_index()
 .rename_axis(columns=None)
)

输出：

  customer_type    Process  2022-01-01  2022-01-02  2022-01-03
0             A  process_A           2           4           0
1             A  process_B           3           9           0
2             B  process_A           0           0           5
3             B  process_B           0           0           6