How to Use assign in pandas: Add or Modify Columns Easily

Use the assign method in pandas to add new columns or modify existing ones in a DataFrame. It returns a new DataFrame with the changes, leaving the original unchanged. You pass column names as keywords and their values as expressions or functions.

Syntax

The assign method syntax is:

• DataFrame.assign(**kwargs)

Here, kwargs are column names as keys and their new values or expressions as values. It returns a new DataFrame with these columns added or updated.

python

df.assign(new_col=value_expression)

Example

This example shows how to add a new column and modify an existing one using assign. The original DataFrame stays the same.

python

import pandas as pd

df = pd.DataFrame({
    'A': [1, 2, 3],
    'B': [4, 5, 6]
})

# Add new column 'C' as sum of 'A' and 'B', and modify 'B' by doubling
new_df = df.assign(C=df['A'] + df['B'], B=lambda x: x['B'] * 2)

print("Original DataFrame:\n", df)
print("\nNew DataFrame with assign:\n", new_df)

Output

Original DataFrame:
    A  B
0  1  4
1  2  5
2  3  6

New DataFrame with assign:
    A   B  C
0  1   8  5
1  2  10  7
2  3  12  9

Common Pitfalls

Common mistakes when using assign include:

• Expecting assign to modify the original DataFrame (it returns a new one).
• Passing positional arguments instead of keyword arguments.
• Using column names that conflict with existing DataFrame methods.

Always assign the result back to a variable if you want to keep the changes.

python

import pandas as pd

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

# Wrong: This does not change df
# df.assign(B=[3, 4])

# Right: Assign result to a new variable or overwrite
new_df = df.assign(B=[3, 4])
print(new_df)

Output

   A  B
0  1  3
1  2  4

Quick Reference

Summary tips for using assign:

• Use keyword arguments to add or update columns.
• Values can be scalars, lists, Series, or functions.
• Functions receive the DataFrame and return the new column values.
• Original DataFrame is not changed; assign returns a new one.

Key Takeaways

The assign method adds or modifies columns and returns a new DataFrame.

Always assign the result of assign to a variable to keep changes.

You can pass functions to assign to compute new column values dynamically.

Original DataFrame remains unchanged after using assign.

Use keyword arguments with column names and values or functions.