Hi all, I’m using pymc3 to do some Bayesian A/B testing. The beauty of this approach – in contrast to traditional frequentist approaches – is that there are a variety of metrics you can generate and learn a hell of a lot more about our experiment. I’m most interested in computing the error metrics…

The probability that \lambda_B > \lambda_A (section 6.1) can be calculated (approximately) directly: pBgtA = (trace.posterior['Control'] > trace.posterior['Optimisation']).mean() The expected loss with would be something like this: upliftLossA = (trace.posterior['Control'] - trace.posterior['Opti…

Creating custom Joint Distribution in pymc3

Questions

cluhmann May 10, 2022, 6:02pm 2

Welcome!

How did you sample the posteriors you have?

Topic		Replies	Views
Unexpected results Questions	2	507	February 15, 2019
How to set up a custom likelihood function for two variables Questions	8	13306	March 9, 2018
Create Custom Joint Distribution Questions	9	2610	May 26, 2019
Fitting joint distributions using custom distribution as part of mixture Questions	3	515	January 13, 2020
Joint PDF Surrogate Model Questions	9	1696	July 30, 2018

Creating custom Joint Distribution in pymc3

Related topics