在Python中使用Pivot_Table的问题：是否有任何方法可以保持数据的原始顺序并且没有多索引？

Question

我正在尝试重新创建数据框。以下是原始数据框：

df = pd.DataFrame([['January','Monday',0,1,20],['January','Monday',1,2,15],['January','Wednesday',0,1,35],['March','Monday',0,1,23],['March','Monday',1,2,50],['March','Monday',2,3,60] ,['April','Wednesday',0,1,75]],columns = ['Month','Day','Data1','Data2','Random'])

     Month        Day  Data1  Data2  Random
0  January     Monday      0      1      20
1  January     Monday      1      2      15
2  January  Wednesday      0      1      35
3    March     Monday      0      1      23
4    March     Monday      1      2      50
5    March     Monday      2      3      60
6    April  Wednesday      0      1      75

我的目标是实现以下结果：

     Month        Day  0    1    2
0  January     Monday  1  2.0  NaN
1  January     Monday  1  2.0  NaN
2  January  Wednesday  1  NaN  NaN
3    March     Monday  1  2.0  3.0

我尝试如下使用pivot_table，但是它当然不起作用，因为pivot_table不允许索引重复，而且我还将拥有multiindex，这会在以后的过程中引起问题。

df1 = pd.pivot_table(df, values = 'Data2', index = ['Month','Day'], columns = ['Data1'])

Data1                0    1    2
Month   Day                     
April   Wednesday  1.0  NaN  NaN
January Monday     1.0  2.0  NaN
        Wednesday  1.0  NaN  NaN
March   Monday     1.0  2.0  3.0

还有其他方法可以得到我想要的结果吗？提前非常感谢。

Answer 1

您可以尝试使用unby的groupby：

df.groupby(['Month','Day','Data2'])['Data2'].first().unstack().reset_index()

输出：

Data2    Month        Day    1    2    3
0        April  Wednesday  1.0  NaN  NaN
1      January     Monday  1.0  2.0  NaN
2      January  Wednesday  1.0  NaN  NaN
3        March     Monday  1.0  2.0  3.0