CODEWITHSUNDEEP
pythonpandas
Biswankar Das
Biswankar Das

Reputation: 305

How to set multilevel index in a dataframe with existing columns

I have a dataframe as below:

     A      B     C     D
0    hola   32    43    54
1    hey    87    67    45
2    hi     10    54    89
3    hola   19    34    12
4    hi     11    59    09

I need to set a multilevel index using A and B , which is grouped by A I need the following dataframe

 A      B     C     D
hola   32    43    54
       19    34    12
hey    87    67    45
hi     10    54    89
       11    59    09

I have tried using df.set_index(['A','B']) and i get 

A      B     C     D
hola   32    43    54
hola   19    34    12
hey    87    67    45
hi     10    54    89
hi     11    59    09

Upvotes: 2

Views: 9520

Answers (2)

Andrew L
Andrew L

Reputation: 7038

Sorting first is not necessary- you are setting the MultiIndex but it is not lexsorted:

df.set_index(['A','B']).index.is_lexsorted()
False

As an alternate method, set the index as you already have and sort it after the fact:

df.set_index(['A','B']).sort_index()
          C   D
A    B         
hey  87  67  45
hi   10  54  89
     11  59   9
hola 19  34  12
     32  43  54

Upvotes: 0

Allen Qin
Allen Qin

Reputation: 19957

You need to sort first.

df.sort_values(['A','B']).set_index(['A','B'])
Out[60]: 
          C   D
A    B         
hey  87  67  45
hi   10  54  89
     11  59   9
hola 19  34  12
     32  43  54

Upvotes: 5

Related Questions