如何删除pandas数据透视表中的多级索引

Question

我有一个给定的数据帧:

df = {'TYPE' : pd.Series(['Advisory','Advisory1','Advisory2','Advisory3']),
 'CNTRY' : pd.Series(['IND','FRN','IND','FRN']),
 'VALUE' : pd.Series([1., 2., 3., 4.])}
df = pd.DataFrame(df)
df = pd.pivot_table(df,index=["CNTRY"],columns=["TYPE"]).reset_index()

在旋转之后,我如何获得具有列的数据框并且df如下所示; 删除多级索引,VALUE

Type|CNTRY|Advisory|Advisory1|Advisory2|Advisory3
0     FRN     NaN      2.0      NaN     4.0 
1     IND     1.0      NaN      3.0     NaN 

Answer 1

你可以添加参数values:

df = pd.pivot_table(df,index="CNTRY",columns="TYPE", values='VALUE').reset_index()
print (df)
TYPE CNTRY  Advisory  Advisory1  Advisory2  Advisory3
0      FRN       NaN        2.0        NaN        4.0
1      IND       1.0        NaN        3.0        NaN

并删除列名称rename_axis:

df = pd.pivot_table(df,index="CNTRY",columns="TYPE", values='VALUE') \
       .reset_index().rename_axis(None, axis=1)
print (df)
  CNTRY  Advisory  Advisory1  Advisory2  Advisory3
0   FRN       NaN        2.0        NaN        4.0
1   IND       1.0        NaN        3.0        NaN

但也许只有必要pivot:

df = df.pivot(index="CNTRY",columns="TYPE", values='VALUE') \
       .reset_index().rename_axis(None, axis=1)
print (df)
  CNTRY  Advisory  Advisory1  Advisory2  Advisory3
0   FRN       NaN        2.0        NaN        4.0
1   IND       1.0        NaN        3.0        NaN

因为pivot_table聚合重复默认聚合函数mean:

df = {'TYPE' : pd.Series(['Advisory','Advisory1','Advisory2','Advisory1']),
 'CNTRY' : pd.Series(['IND','FRN','IND','FRN']),
 'VALUE' : pd.Series([1., 4., 3., 4.])}
df = pd.DataFrame(df)
print (df)
  CNTRY       TYPE  VALUE
0   IND   Advisory    1.0
1   FRN  Advisory1    1.0 <-same FRN and Advisory1 
2   IND  Advisory2    3.0
3   FRN  Advisory1    4.0 <-same FRN and Advisory1 

df = df.pivot_table(index="CNTRY",columns="TYPE", values='VALUE')
       .reset_index().rename_axis(None, axis=1)
print (df)
TYPE   Advisory  Advisory1  Advisory2
CNTRY                                
FRN         0.0        2.5        0.0
IND         1.0        0.0        3.0

替代groupby,聚合函数和unstack:

df = df.groupby(["CNTRY","TYPE"])['VALUE'].mean().unstack(fill_value=0)
      .reset_index().rename_axis(None, axis=1)
print (df)
  CNTRY  Advisory  Advisory1  Advisory2
0   FRN       0.0        2.5        0.0
1   IND       1.0        0.0        3.0