Set index of a multi-index which has been reset Python/pandas-CodePudding

I simply want the reverse operation of this command:

.reset_index(level=['ID'], inplace=True, col_level=1)

Now I have a dataframe:

           DONNES_PRODUITS                        ... PYC_INTEGRES            
        ID             UE              PERIMETRE  ...        TVSU8 TV001 TV002
NUMBER                                            ...                         
74       2             U4A                MAIZON  ...          NaN   NaN   NaN
74       3             U4A                MAIZON  ...          NaN   NaN   NaN
74       6             U4A                MAIZON  ...          NaN   NaN   NaN

I want set ID as an index with NUMBER, for created a multi_index.

expected output:

           DONNES_PRODUITS                        ... PYC_INTEGRES            
                       UE               PERIMETRE ...        TVSU8 TV001 TV002
NUMBER  ID                                        ...                         
74       2             U4A                MAIZON  ...          NaN   NaN   NaN
74       3             U4A                MAIZON  ...          NaN   NaN   NaN
74       6             U4A                MAIZON  ...

I tried this solution:

result.set_index(['ID'], append=True)

I got the error message None of ['ID'] are in the columns

CodePudding user response：

You use .reset_index(level=['ID'], inplace=True, col_level=1) so ID is reset to the level 1 of multi-index columns, you can use

df = df.set_index([('', 'ID')], append=True)

print(df)

              DONNES_PRODUITS
                           UE
NUMBER (, ID)
74     2                  U4A
       3                  U4A
       6                  U4A

df = (df.swaplevel(0,1,1)
      .set_index(['ID'], append=True)
      .swaplevel(0,1,1))

print(df)

          DONNES_PRODUITS
                       UE
NUMBER ID
74     2              U4A
       3              U4A
       6              U4A