Data portfolio

Revenue per Visitor Statistical Significance Calculator

Developed a Shiny app in Python that provides a more reliable method for calculating statistical significance in Revenue Per Visitor (RPV) using the Mann–Whitney U test, since revenue distribution is not a normal distribution, resulting in more accurate results for assessing statistical significance. 

• Below there is the embedded resulting app, but also you can access directly here

• If you are curious and want to check how I did it, feel free to explore the code at my github repository

🚧 Reminder: this calculator has a frequentist approach; work in progress in my todo list is to develop a RPU calculator with a bayesian approach (I'm thinking Gamma distribution, but there are approaches with poisson, lognormal and gaussian, ... also cool to use PYMC package for inference but it takes quite a while... check my jupyter notebooks to see my progress and my attempts).