Pandas: How to divide one DataFrame by another (element-wise)

Updated: February 22, 2024 By: Guest Contributor Post a comment


In data analysis, dividing one DataFrame by another is a common operation, especially in finance and economics, where changes between datasets are frequently examined. Pandas, a powerful data manipulation library in Python, makes this task straightforward with its built-in functionalities. This tutorial introduces you to dividing one DataFrame by another element-wise, using various examples to help you grasp the concept from basic to more advanced scenarios.

Before diving into the examples, ensure you’ve installed Pandas. If not, you can install it using pip:

pip install pandas

Let’s start with the basics.

Basic Division of DataFrames

Create two DataFrames with the same dimensions:

import pandas as pd

# Create DataFrame A
A = pd.DataFrame({
    'A': [10, 20, 30],
    'B': [40, 50, 60]

# Create DataFrame B
B = pd.DataFrame({
    'A': [2, 5, 10],
    'B': [4, 10, 20]

To divide A by B element-wise, we use the divide method:

result = A.divide(B)

The output will be:

     A    B
0  5.0  10.0
1  4.0  5.0
2  3.0  3.0

This demonstrates the simplest form of division where each element of A is divided by the corresponding element in B, resulting in a new DataFrame result.

Handling Mismatched Indices

Often, DataFrames have mismatched indices. Here’s how to handle them:

import pandas as pd

# DataFrames with different indices
A = pd.DataFrame({'A': [10, 20, 30, 40], 'B': [50, 60, 70, 80]}, index=[0, 1, 2, 3])
B = pd.DataFrame({'A': [2, 4], 'B': [5, 10]}, index=[2, 3])

# Fill missing indices in B with 1 (to avoid division by zero)
B_reindexed = B.reindex_like(A).fillna(1)

# Element-wise division
result = A.divide(B_reindexed)

The output showcases how indices that do not match are handled by filling them with a default value, in this case, 1, to ensure a smooth division process:

      A     B
0  10.0  10.0
1  20.0  60.0
2  15.0  10.0
3  20.0   8.0

Dividing DataFrames with Different Shapes

There are scenarios where you need to divide two DataFrames of different shapes. Pandas allows for this through broadcasting, similar to NumPy:

import pandas as pd

A = pd.DataFrame({'A': [10, 20, 30], 'B': [40, 50, 60]})
B = pd.DataFrame({'A': [2, None, 10], 'B': [4, 10, None]}, index=[0, 1, 2])

# Handling None values and broadcasting
B = B.fillna(1)
result = A / B

The fillna(1) method is used to handle None values to avoid division by zero, and the division is performed across the DataFrames, demonstrating the power of Pandas’ broadcasting capability. The output would similarly reflect the division operation element-wise.

Advanced Operations: Division with Conditions

Advanced scenarios might require conditional operations during the division. For instance, dividing based on specific criteria:

import pandas as pd

A = pd.DataFrame({'A': [100, 200, 300], 'B':[400, 500, 600]})
B = pd.DataFrame({'A': [10, 20, 30], 'B': [40, 50, 60]}, index=[0, 1, 2])

# Conditional Division
result = A.div(B.where(B > 15, 1))

This example demonstrates using the where method to apply conditions during division. Data in B less than or equal to 15 is replaced by 1 (to avoid division by zero), and the division is applied. The output illustrates how conditions can influence element-wise operations to achieve specific analysis objectives:

      A    B
0  10.0  10.0
1  10.0  10.0
2  10.0  12.0


Element-wise division of one DataFrame by another in Pandas is a versatile operation, supporting various data processing needs. This tutorial covered basic to advanced scenarios, demonstrating how Pandas efficiently handles these operations with ease. Understanding these methods will significantly enhance your data manipulation skills in Pandas, making your data analysis tasks more dynamic and in-depth.