Pandas >> How to Sort in Pandas

Table of Contents

In this article, we will talk about how to sort data in DataFrame.

Sometimes we want to display data in some order. We can use sort functions of pandas to achieve this. h Firstly, we prepare data for demonstrating the sort function.

Preparing data

import pandas as pd

df = pd.DataFrame({
    "name": ["Kevin", "Jack", "Mary", "Bob", "Robert", "Amy"],
    "score": [80, 90, 95, 93, 88, 81],
    "class": ["A", "B", "A", "A", "B", "B"]
}, index=["K", "J", "M", "B", "R", "A"])

df

Result

Sorting by one column

We can use sort_values method and specify the column name to sort by a single column.
Attention: sort_values method will not change origin DataFrame and it will return a sorted DataFrame.

df.sort_values("score")
# OR
df.sort_values(by="score")

Result

We can also specify the sort order by ascending option.
True is ascending order, False is descending order.

# ascending order
df.sort_values("score", ascending=True)

# descending order
df.sort_values("score", ascending=False)

Sorting by multiple columns

We can specify a list or array of column names to sort by multiple columns.

df.sort_values(by=["class", "score"])

Result

We can use ascending option to specify sort order for all columns.

df.sort_values(by=["class", "score"], ascending=False)

Result

We can also specify sort order for each column by specifying a list or array of ascending boolean values with same length.

df.sort_values(by=["class", "score"], ascending=[False, True])

Result

Sorting by Index

We can use sort_index method to sort data in DataFrame. This method also don’t change original DataFrame and return the sorted DataFrame.

df.sort_index()
df.sort_index(ascending=False)

If we want to sort in columns direction, we can specify axis option.

df.sort_index(axis="columns")
# OR 
df.sort_index(axis="columns", ascending=True)
# OR 
df.sort_index(axis=1)
# OR 
df.sort_index(axis=1, ascending=True)

Result