E-Commerce Data Analysis with Pandas

Project Description

In this project, you will work with the UCI Online Retail Dataset, a real transactions dataset from a UK-based online store with over 500,000 rows. You will clean it, filter it, and compute your first business metric.

This project is closer to real work than most beginner datasets. The data is not clean, and some rows make no sense: you have to deal with that before doing any analysis.

Project Requirements

• Download the UCI Online Retail Dataset (available on Kaggle or the UCI ML Repository)

• Sample 10% of rows to keep things manageable

• Clean the data: remove nulls, fix data types, filter out returns, and free items

• Convert InvoiceDate to a proper datetime object

• Create a Revenue column (Quantity × UnitPrice)

• Find the top 10 countries by total revenue and plot the result

Technologies to Use

• Python

• Pandas

• Matplotlib / Seaborn

• Jupyter Notebook

What You Will Learn

You will practice cleaning a large, realistic dataset and computing a derived metric. You will also understand why negative quantities and zero prices exist in real transaction data, and how to handle them without deleting useful rows.

Want to See a Solution?

A full walkthrough of this project is available on Towards Data Science: 🔗 EDA in Public: Cleaning and Exploring Sales Data with Pandas