Transaction Pattern Analysis

A Data-Driven Analysis of Temporal Sales Dynamics and Demand Peaks

View Notebook · Dataset · Results · Contributing

---

Overview

This project investigates temporal purchasing behavior within a retail bakery environment using transactional sales data. The objective is to identify patterns in customer demand across time and evaluate how these patterns can inform operational decision-making.

The analysis provides insight into how temporal dynamics influence retail performance and demonstrates how data-driven approaches can support optimization of:

• Inventory management
• Staffing allocation
• Production scheduling

---

Features

Category	Techniques
Data Preprocessing	Cleaning, type conversion, missing value handling
Feature Engineering	Time-based feature extraction (hourly segmentation)
Exploratory Analysis	Distribution analysis, correlation matrices, trend visualization
Statistical Testing	Relationship analysis between price, quantity, and time
Clustering	K-Means (manual implementation + scikit-learn)
Classification	K-NN (manual implementation + scikit-learn)
Dimensionality Reduction	Principal Component Analysis (PCA)

---

Project Structure

transaction-pattern-analysis/
├── transaction_pattern_analysis.ipynb  # Main analysis notebook
├── Bakery sales.csv                    # Raw dataset (234,005 transactions)
├── Bakery_Sales1.json.zip              # JSON export
└── README.md                           # Project documentation

---

Dataset

The dataset contains 234,005 transactional records from a French retail bakery spanning from January 2021 to September 2022.

Field	Description
date	Transaction date
time	Transaction time (HH:MM)
ticket_number	Unique transaction identifier
article	Product name
Quantity	Number of items purchased
unit_price	Price per unit (€)

Source: Kaggle - French Bakery Daily Sales

Note: Data cleaning is performed directly within the notebook to ensure transparency and reproducibility.

---

Results

Sales Distribution by Hour

The analysis reveals clear peak demand periods throughout the day, enabling targeted staffing and inventory decisions.

Peak Hours: Morning rush (8-10 AM) and afternoon (12-2 PM)

Clustering Analysis

K-Means clustering segments transactions into distinct behavioral groups based on quantity and pricing patterns.

PCA Insights

Principal Component Analysis reduces dimensionality while preserving variance, revealing underlying structure in transaction data.

---

Tech Stack

Tool	Purpose
	Core programming language
	Data manipulation & analysis
	Numerical computing
	Data visualization
	Statistical visualization
	Machine learning

---

Getting Started

Prerequisites

• Python 3.8+
• Jupyter Notebook or Google Colab

Installation

1. Clone the repository

   git clone https://github.com/anaya33/transaction-pattern-analysis.git
   cd transaction-pattern-analysis

2. Install dependencies

   pip install pandas numpy matplotlib seaborn scikit-learn

3. Launch the notebook

   jupyter notebook transaction_pattern_analysis.ipynb

Using Google Colab

1. Open Google Colab
2. Upload transaction_pattern_analysis.ipynb
3. Upload Bakery sales.csv or connect via GitHub
4. Run cells sequentially

---

Contributing

Contributions are welcome! Here's how you can help:

Ways to Contribute

• Bug Reports — Found an issue? Open a detailed bug report
• Feature Requests — Have ideas for new analyses? Share them!
• New Visualizations — Add compelling charts or dashboards
• Additional ML Models — Implement other clustering/classification algorithms
• Documentation — Improve explanations or add tutorials
• Code Optimization — Enhance performance or code quality

Contribution Guidelines

1. Fork the repository
2. Create a feature branch

   git checkout -b feature/your-feature-name

3. Commit your changes with clear messages

   git commit -m "Add: description of your changes"

4. Push to your branch

   git push origin feature/your-feature-name

5. Open a Pull Request with a detailed description

Ideas for Contributors

Difficulty	Task
Easy	Add more visualizations (box plots, violin plots)
Easy	Improve code comments and documentation
Medium	Implement DBSCAN or hierarchical clustering
Medium	Add time series forecasting (ARIMA, Prophet)
Medium	Create an interactive dashboard (Plotly/Dash)
Advanced	Build a recommendation system for products
Advanced	Deploy as a web application

---

License

This project is licensed under the MIT License — see the LICENSE file for details.

---

Acknowledgments

• Dataset provided by Matthieu Gimbert on Kaggle
• Developed as part of Advanced Data Analytics (ITEC 4220) coursework

---

Star this repo if you found it helpful!

View Live Project · Download JSON Data