How to Sort DataFrame in pandas: Simple Guide

To sort a DataFrame in pandas, use df.sort_values() to sort by column values or df.sort_index() to sort by the index. You can specify ascending or descending order and sort by multiple columns easily.

Syntax

The main method to sort a DataFrame by column values is df.sort_values(by=column_name, ascending=True). Use by to specify the column(s) to sort by. Set ascending=False to sort in descending order. To sort by index, use df.sort_index(ascending=True).

python

df.sort_values(by='column_name', ascending=True)
df.sort_index(ascending=True)

Example

This example shows how to sort a DataFrame by one column in ascending order and by multiple columns with different orders.

python

import pandas as pd

data = {'Name': ['Alice', 'Bob', 'Charlie', 'David'],
        'Age': [25, 30, 25, 35],
        'Score': [85, 90, 85, 80]}
df = pd.DataFrame(data)

# Sort by Age ascending
sorted_df = df.sort_values(by='Age')

# Sort by Age ascending, then Score descending
sorted_multi = df.sort_values(by=['Age', 'Score'], ascending=[True, False])

print('Sorted by Age:')
print(sorted_df)
print('\nSorted by Age and Score:')
print(sorted_multi)

Output

Sorted by Age:
      Name  Age  Score
0    Alice   25     85
2  Charlie   25     85
1      Bob   30     90
3    David   35     80

Sorted by Age and Score:
      Name  Age  Score
0    Alice   25     85
2  Charlie   25     85
1      Bob   30     90
3    David   35     80

Common Pitfalls

• Forgetting to assign the sorted DataFrame back to a variable or use inplace=True means the original DataFrame stays unsorted.
• Using ascending as a single boolean when sorting by multiple columns causes unexpected results; use a list of booleans instead.
• Sorting by a column name that does not exist will raise an error.

python

import pandas as pd

data = {'A': [3, 1, 2]}
df = pd.DataFrame(data)

# Wrong: does not change df
sorted_wrong = df.sort_values(by='A')
print(df)  # Original order

# Right: assign or use inplace
sorted_right = df.sort_values(by='A')
print(sorted_right)

# Or inplace
# df.sort_values(by='A', inplace=True)
# print(df)

Output

   A
0  3
1  1
2  2
   A
1  1
2  2
0  3

Quick Reference

Method	Purpose	Key Parameters
sort_values(by, ascending=True, inplace=False)	Sort by column values	by: column(s) name(s), ascending: bool or list, inplace: bool
sort_index(ascending=True, inplace=False)	Sort by index	ascending: bool, inplace: bool

Key Takeaways

Use df.sort_values() to sort by one or more columns with control over ascending or descending order.

Remember to assign the result or use inplace=True to keep the sorted DataFrame.

Use df.sort_index() to sort rows by their index labels.

When sorting by multiple columns, pass a list of booleans to ascending to control each column's order.

Sorting by a non-existent column will cause an error, so check column names carefully.