.sort_values()

In Pandas, the .sort_values() function is used to sort values in a DataFrame by one or more columns. This function is useful in data analysis, data visualization, data cleaning, and more.

Syntax

DataFrame.sort_values(by, axis=0, ascending=True, inplace=False, kind='quicksort', na_position='last', ignore_index=False, key=None)

Parameters:

• by: A label or a list of labels to sort by (levels can also be specified if applicable).
• axis: Specifies which axis will be sorted (0 or index, 1 or columns); defaults to 0.
• ascending: Specifies whether the sort will be ascending or descending (True or False); defaults to True.
• inplace: By setting it to True, the operation will be performed on the original DataFrame and the function will return None; defaults to False.
• kind: Specifies which algorithm to use: 'quicksort', 'mergesort', 'heapsort', or 'stable'; defaults to 'quicksort'.
• na_position: Specifies where to put the NaN values; defaults to last.
• ignore_index: Ignores the original index and put a new ascending set of values in its place. The default is False.
• key: Applies a function before sorting.

Return value:

The .sort_values() function returns a sorted DataFrame (or None if inplace=True).

Example 1: Sort by a Single Column

In this example, the DataFrame is sorted in ascending order based on the Score column:

import pandas as pd

data = {
  'Name': ['Alice', 'Bob', 'Charlie', 'David'],
  'Score': [88, 92, 85, 90]
}

df = pd.DataFrame(data)

sorted_df = df.sort_values(by='Score')

print(sorted_df)

Copy to clipboard
Copy to clipboard

Here is the output:

      Name  Score
2  Charlie     85
0    Alice     88
3    David     90
1      Bob     92

Copy to clipboard
Copy to clipboard

Example 2: Sort by Multiple Columns

In this example, the data is first sorted by Department, and then by Score within each department:

import pandas as pd

data = {
  'Name': ['Alice', 'Bob', 'Charlie', 'David', 'Eve'],
  'Department': ['HR', 'Finance', 'HR', 'Finance', 'HR'],
  'Score': [88, 92, 85, 90, 85]
}

df = pd.DataFrame(data)

sorted_df = df.sort_values(by=['Department', 'Score'])

print(sorted_df)

Copy to clipboard
Copy to clipboard

Here is the output:

      Name Department  Score
3    David    Finance     90
1      Bob    Finance     92
2  Charlie         HR     85
4      Eve         HR     85
0    Alice         HR     88

Copy to clipboard
Copy to clipboard

Codebyte Example: Sort in Descending Order with NaNs First

In this codebyte example, missing values (NaNs) appear at the top, and the Price column is sorted in descending order:

Visit us
Visit us
Hide code
Code
Output
Hide output
Loading...
Copy to your clipboard
Copy Codebyte
Run

Frequently Asked Questions

1. What’s the difference between .sort_values() and .sort_index()?

.sort_values() sorts by column values, while .sort_index() sorts by row or column index. Use .sort_index() when sorting based on DataFrame labels rather than content.

2. Does .sort_values() modify the original DataFrame?

By default, .sort_values() returns a new DataFrame. To modify the original, set inplace=True.

3. Can I sort a Series using .sort_values()?

Yes. Pandas Series also supports .sort_values(), and the syntax is simpler since there’s no need to specify a column.