Data Cleaning & Preprocessing for Superstore Sales Data

Project Overview

Cleaned and prepared a Superstore sales dataset using Python and Pandas to make the data reliable and ready for further analysis.

Data Cleaning & Validation

* Inspected the dataset structure, columns, data types, and summary statistics.

* Checked for missing values and handled missing Postal Code values using the mode.

* Checked for duplicate records.

* Validated and converted date columns where necessary.

* Removed leading and trailing spaces from text fields.

* Checked categorical columns for inconsistent formatting.

* Examined numerical variables for potential outliers using summary statistics, boxplots, and the IQR method.

Outlier Analysis

Outliers were investigated in Sales, Profit, Quantity, and Discount using the IQR method.

After reviewing the detected values, the outliers were retained because they represented valid business transactions rather than data-entry errors.

Feature Engineering

Created additional features from the existing dataset, including:

* Profit Margin

* Shipping Duration

* Sales Performance Category

Reusable Data Processing

Developed a reusable Object-Oriented `DataAnalyzer` framework to organize preprocessing and feature-engineering operations and make the workflow easier to reuse on similar datasets.

Memory Optimization

Optimized the dataset's memory usage by converting suitable numerical and categorical columns to more efficient data types.

Tools & Technologies

Python, Pandas, NumPy, Matplotlib, Seaborn, Jupyter Notebook