RadvancedaiAI generated

Bayesian A/B Testing Framework

Build an R set of functions that runs Bayesian A/B test analysis on conversion data, computing posterior distributions and probability that variant B beats A. Learners practice Bayesian inference concepts, Beta-Binomial conjugate modeling, and building reusable statistical functions.

Estimate
~8.5h
Steps
5
Completed by
0
Proposed by
codeseed.app

stats · ggplot2

Project roadmap

  1. 01

    Model conversions with Beta-Binomial priors

    ~2h

    Write a function that takes conversion counts for two variants and computes posterior Beta distributions.

  2. 02

    Simulate posterior samples

    ~1.5h

    Draw random samples from each posterior and estimate the probability that variant B outperforms A.

  3. 03

    Compute expected loss and credible intervals

    ~2h

    Calculate expected loss for choosing each variant and 95% credible intervals for the conversion rate difference.

  4. 04

    Visualize posterior distributions

    ~1.5h

    Plot overlapping posterior density curves for both variants with the probability-of-improvement annotated.

  5. 05

    Package as reusable functions

    ~1.5h

    Organize the functions into a small package-like structure with documentation and example usage.

Resources

Ready to build this?

Get a GitHub repo and start building. Your AI reviewer checks each step as you go.

~8.5h · 5 steps

Tech stack

statsggplot2