🧪 A/B Testing Simulator

An interactive A/B testing simulator and analysis tool built with Streamlit. This app enables users to generate rich synthetic data, apply various frequentist statistical tests as well as advanced Bayesian A/B testing, calculate required sample sizes, and visualize results.

---

🚀 Features

• 🧠 Simulate User Behavior
  Generate synthetic A/B test datasets with rich features (e.g., device, traffic source, session duration, signup month).

• 📊 Statistical Tests Supported

  • Z-test for proportions
  • Welch’s t-test for continuous metrics
  • Chi-squared test for feature relationships
  • Bayesian A/B test (Beta posteriors)

• 📈 Power & Sample Size Calculator
  Estimate the number of users required to detect a specific uplift.

• 🎨 Streamlit Dashboard
  Easily explore results with visualizations and summary tables.

---

🛠️ Project Structure

ab_testing_simulator/
│
├── app/
│ ├── init.py
│ ├── main.py # Streamlit app entrypoint
│ ├── data_generator.py # Simulates A/B test data
│ ├── utils.py # Summary, sample size, helper functions
│ └── tests.py # Statistical testing functions
│
├── .gitignore
├── requirements.txt
├── streamlit_app.py # Optional Streamlit entry file (used in deployment)
└── README.md

▶️ Getting Started

1. Clone the Repository

git clone https://github.com/your-username/AB-Testing-Simulator.git
cd AB-Testing-Simulator

2. Install Dependencies

(Recommended: use a virtual environment)

pip install -r requirements.txt

3. Run the App

streamlit run app/main.py

---

📦 Requirements

Python 3.8+ Streamlit NumPy, Pandas, SciPy, Statsmodels, Faker, Matplotlib All dependencies are listed in requirements.txt.

🧠 Use Cases

• Learn and practice A/B testing
• Demonstrate statistical decision-making
• Simulate realistic experiments without real data
• Teaching tool for marketing, product, or data teams