How to sort values within a column of a pandas dataframe ?

Pandas Dataframes are widely used in data analysis and manipulation. Sorting the values within a column of a pandas dataframe is one of the most common tasks while working with data. It can be achieved by using the sort_values() function. Examples:

Sorting the values within a column

The syntax for this method is as follows:

df['column name'].sort_values(ascending=True or False)

where, 'column name' corresponds to the column in the dataframe which needs to be sorted. The ascending parameter is used to specify whether the values need to be sorted in ascending (smallest to largest) or descending (largest to smallest) order. It is set to True by default.

Sort the values within a column in ascending order

For example, if you have a dataframe df that contains the following values:

import pandas as pd

data = {'User_id':[0,1,2,3,4], 
        'Age':[25,30,29,26,22],
        'Name':['Ben','Anna','Zoe','Tom','John']}

df = pd.DataFrame(data)

Output

       User_id  Age  Name
    0        0   25   Ben
    1        1   30  Anna
    2        2   29   Zoe
    3        3   26   Tom
    4        4   22  John

To sort the values in the 'Age' column in ascending order, run the following command:

df.sort_values('Age', ascending = True)

This would give an output as :

       User_id  Age  Name
    4        4   22  John
    0        0   25   Ben
    3        3   26   Tom
    2        2   29   Zoe
    1        1   30  Anna

Note that

df.sort_values('Age')

returns the same output.

Sort the values within a column in descending order

Similarly, to sort in descending order, set the parameter ascending = False:

df.sort_values('Age', ascending = False)

This would give an output as :

       User_id  Age  Name
    1        1   30  Anna
    2        2   29   Zoe
    3        3   26   Tom
    0        0   25   Ben
    4        4   22  John

Additional features

Save sorted output

For those seeking to permanently modify the dataframe, an additional feature is available. This allows you to save your sorted output and make it a permanent alteration in the dataframe structure, a solution is to add option inplace=True:

df.sort_values(by=['Age'], inplace=True)

Show sorted output for a given column

To only show the Age column:

 df['Age'].sort_values()

Output

    4    22
    0    25
    3    26
    2    29
    1    30
    Name: Age, dtype: int64

Sort the values within a column of strings

Another example sorting the dataframe using the 'Name' column

df.sort_values(by=['Name'])

print(df.sort_values(by=['Name']))

returns

        User_id  Age  Name
    1        1   30  Anna
    0        0   25   Ben
    4        4   22  John
    3        3   26   Tom
    2        2   29   Zoe

Sort by multiple columns in the dataframe

You can also sort by multiple columns in the dataframe. To do so, you specify a list of column names as an argument to the sort_values() function. The syntax for this is:

df.sort_values(['column1', 'column2'], ascending = True)

where, column1 and column2 correspond to the columns in the dataframe which need to be sorted. This will sort column 1 values in ascending order followed by column 2 values in ascending order.

For example, if the dataframe contains a column 'Gender' in addition to 'Age', you can sort values in both columns together by running:

import pandas as pd

data = {'User_id':[0,1,2,3,4], 
        'Age':[20,20,20,26,22],
        'Name':['Ben','Anna','Zoe','Tom','John']}

df = pd.DataFrame(data)

Output

       User_id  Age  Name
    0        0   20   Ben
    1        1   20  Anna
    2        2   20   Zoe
    3        3   26   Tom
    4        4   22  John

then

 df.sort_values(by=['Age','Name'])

This would give an output as :

       User_id  Age  Name
    1        1   20  Anna
    0        0   20   Ben
    2        2   20   Zoe
    4        4   22  John
    3        3   26   Tom

References