Bayesian ab testing python github
aByes is a Python package for Bayesian A/B Testing, which supports two main decision rules: A lot of the underlying theory is discussed in this blog post. In your target
folder, clone the repository with the command: Then, inside the same folder (as always, it is advisable to use a virtual environment): To check that the package has been installed, in the Python shell type: If everything works correctly, the package will be imported without errors. Dependencies
information). How to use aByesThe main steps to run the analysis of an A/B experiment are:
These and many more examples and instructions can be found in this blogpost. Example
LimitationsCurrently, aByes:
These shortcomings may be improved in future versions of aByes. (Feel free to fork the project and make these improvements yourself!) LicenceApache License, Version 2.0 Bayesian Tests are a Bayesian alternative to classical hypothesis testing, you can read more about it
here. Simply install from Suppose you run an AB test between two populations, the outcome being one set of measures for each group (could be conversion, revenue, latency, etc). We'll call these two data sets You simply need to create a
This will typically produce the following plots: Data and modelThe metric distribution is shown for bothcontrol and variant sets (red bars), as well as a few plots of distributions estimated in the Monte Carlo sampling (blue lines).In our example, one can eyeball that the model is a good fit for the data. Posterior knowledge of distribution parametersFor each of the sets (control , variant ), distribution parameters are estimated. This "knowledge" of parameters is described as a likelihood, or
in other words denstity of credibility, which is represented in the above graphs. In our example, it can be seen that the parameters of the two sets are extremely likely to be different, as their distributions are shifted and
One could conclude that A and B groups are different here. TroubleshootingMacOSInstalling requirementsIn order to install the Matplotlib runtime errorYou may run into this
Simply create a
Requirements
|