In this article, we will discuss how to sort a Pandas dataframe by both index and columns.
Sort DataFrame based on Index
We can sort a Pandas DataFrame based on Index and column using sort_index method.
- To sort the DataFrame based on the index we need to pass axis=0 as a parameter to sort_index method.
- To sort the DataFrame based on the column name then we need to pass axis=1 as a parameter to sort_index method.
Syntax
DataFrame_Name.sort_index(axis=0, ascending=True, inplace=False, kind=’quicksort’)
Parameters
- axis- Specifies on which basis to sort whether based on index or column. By default, it sorts based on index i.e. axis=0.
- ascending- Specifies on which order to sort whether in ascending or descending order. It accepts True or False. By default it is True.
- inplace- It specifies that the changes to the DataFrame is Temporary or Permanent. inplace=False indicates temporary and True indicates permanent.
- kind- Specifies which sorting algorithm to use. It accepts quicksort, mergesort, heapsort, etc. By default, it is quicksort if not mentioned.
DataFrame
|
|
col2 |
col1 |
|---|---|---|
|
3 |
150 |
70 |
|
2 |
170 |
55 |
|
1 |
160 |
60 |
This was the DataFrame we are using in the below codes for sorting. By specifying axis=0 in sort_index method we can sort the DataFrame. Even if we don’t specify axis parameter in sort_index by default it sorts the DataFrame based on row.
Example:
In this example, the DataFrame is a sorted DataFrame based on index labels and this was temporarily sorted in python language.
Python3
# import necessary packagesimport pandas as pd# create 2 dataframes with different indexeshostelCandidates1 = pd.DataFrame({'col2': [150, 170, 160], 'col1': [70, 55, 60]}, index=[3, 2, 1])print('original DataFrame')print(hostelCandidates1)# sorted temporarily based on index labelsprint('Sorted by index')hostelCandidates1.sort_index(axis=0) |
Output
original DataFrame
col2 col1
3 150 70
2 170 55
1 160 60
Sorted by index
col2 col1
1 160 60
2 170 55
3 150 70
Sort DataFrame based on columns
To sort the data in the DataFrame on the basis of column names, then we need to pass axis=1 as a parameter to sort_index method in python language.
Python3
# import necessary packagesimport pandas as pd# create 2 dataframes with different indexeshostelCandidates1 = pd.DataFrame({'col2': [150, 170, 160], 'col1': [70, 55, 60]}, index=[3, 2, 1])print('original DataFrame')print(hostelCandidates1)# sorted temporarily based on column labelsprint('Sorted by column name')hostelCandidates1.sort_index(axis=1) |
Output
original DataFrame
col2 col1
3 150 70
2 170 55
1 160 60
Sorted by column name
col1 col2
3 70 150
2 55 170
1 60 160
Example Code to sort the DataFrame & save changes permanently on original DataFrame:
Here the sort operation is performed directly on the original DataFrame and changes are saved permanently because of inplace=True argument.
Python3
# import necessary packagesimport pandas as pd# create 2 dataframes with different indexeshostelCandidates1 = pd.DataFrame({'col2': [150, 170, 160], 'col1': [70, 55, 60]}, index=[3, 2, 1])print('original DataFrame')print(hostelCandidates1)# sorted permanently based on column labelshostelCandidates1.sort_index(axis=1, inplace=True)print('Modified Original DataFrame')print(hostelCandidates1) |
Output
original DataFrame
col2 col1
3 150 70
2 170 55
1 160 60
Modified Original DataFrame
col1 col2
3 70 150
2 55 170
1 60 160
