📈 Mastering idxmax() in Pandas with Real-World Examples

When analyzing data in Python using pandas, it's common to need the index (or label) of the row or column where the maximum value occurs. That’s exactly where idxmax() shines.

Let’s explore how idxmax() works and walk through several real-life examples where it becomes a powerful tool in your data analysis toolkit.

🧠 What Does idxmax() Do?

The idxmax() function returns the index (row or column label) of the first occurrence of the maximum value along a given axis.

Syntax:

python

CopyEdit

DataFrame.idxmax(axis=0, skipna=True)

• axis=0 (default): returns the index of the max value for each column

• axis=1: returns the column label of the max value for each row

• skipna=True: ignores NaNs

🔧 Setup: Sample Dataset

We'll use a few real-world inspired datasets.

import pandas as pd

# Example 1: Sales data for a week

data = {

'Monday': [200, 300, 250],

'Tuesday': [220, 340, 280],

'Wednesday': [210, 330, 300],

'Thursday': [230, 310, 320],

'Friday': [240, 360, 290]

}

df = pd.DataFrame(data, index=['Store A', 'Store B', 'Store C'])

print(df)

✅ Example 1: Which store had the highest sales each day?

df.idxmax()

📌 Result: Returns which store had the max sales for each day.

✅ Example 2: On which day did each store have their best performance?

df.idxmax(axis=1)

📌 Result: Returns the day each store achieved its highest sales.

📊 Real-World Example 4: Best Performing Marketing Channel

# Conversion data from different channels

conversions = pd.DataFrame({

'Google Ads': [120, 130, 150],

'Facebook Ads': [100, 140, 160],

'Email Campaign': [90, 120, 170]

}, index=['January', 'February', 'March'])

# Which channel performed best each month?

conversions.idxmax(axis=1)

📌 Great for marketers comparing campaigns by time period.

Combine with loc for Full Row Access

✅ Example 3:Find row with highest sales on Friday

max_store_friday = df['Friday'].idxmax()

df.loc[max_store_friday]

📌 This gives you the entire row (i.e., the full week's sales) for the store with the highest sales on Friday.

🏦 Example 4: Highest Revenue Product by Region

revenue = pd.DataFrame({

'North': [50000, 40000, 45000],

'South': [55000, 42000, 47000],

'East': [53000, 43000, 48000],

'West': [52000, 41000, 46000]

}, index=['Product A', 'Product B', 'Product C'])

# Which region has the highest revenue for each product?

revenue.idxmax(axis=1)

📌 Helps in identifying strong regional markets per product.

🚨 Gotchas with idxmax()

• Multiple max values? idxmax() only returns the first occurrence.

• NaNs? Handled by default with skipna=True, but be aware if you disable it.

• If all values are NaN, it raises a ValueError.

✨ Summary

idxmax() is simple but immensely useful in real data analysis tasks — from finding top-performing regions, stores, products, to summarizing peaks across time or groups. It’s an essential part of any pandas analyst’s toolkit.

Kishore Babu Valluri

Senior Data Scientist | Freelance Consultant | AI/ML & GenAI Expert

With deep expertise in machine learning, artificial intelligence, and Generative AI, I work as a Senior Data Scientist, freelance consultant, and AI agent developer. I help businesses unlock value through intelligent automation, predictive modeling, and cutting-edge AI solutions.