bayes-sizer 1.0.0

A Bayesian sample size calculator based on VWO SmartStats

BayesSizer

BayesSizer is a Bayesian sample size calculator inspired by the VWO SmartStats methodology. It helps data scientists and experimenters determine the required sample size for A/B tests using Bayesian inference.

Features

• Computes Bayesian-based sample sizes for A/B testing
• Supports configurable Type I (α) and Type II (β) error rates
• Designed for data science and experimentation teams
• Lightweight, simple, and efficient

Installation

Install BayesSizer from PyPI:

pip install bayessizer

Usage

from bayessizer import bayesian_sample_size

# Example: Calculate required sample size with default parameters
sample_size = bayesian_sample_size(alpha=0.05, beta=0.2, effect_size=0.1, std_dev=1)
print(f"Required sample size per group: {sample_size}")

Parameters

• alpha (float): Type I error rate (default: 0.05)
• beta (float): Type II error rate (default: 0.2)
• effect_size (float): Minimum detectable effect size (default: 0.1)
• std_dev (float): Standard deviation of the population (default: 1)

Example Output

Required sample size per group: 385

Running Tests

To ensure correctness, run:

pytest tests/

Contributing

We welcome contributions! Feel free to submit issues or pull requests.

License

This project is licensed under the MIT License.