FAQ

  • How can I cite uravu in my publication?

    The uravu package has been published in the Journal of Open Source Software, so the following reference should be included where uravu is used in a publication: “A. R. McCluskey & T. Snow, (2020). uravu: Making Bayesian modelling easy(er). Journal of Open Source Software, 5(50), 2214, DOI: 10.21105/joss.02214

  • How do I use the uravu.utils.bayes_factor() function to compare different models?

    The uravu.utils.bayes_factor() function uses the the second Table on page 777 of Kass and Raftery’s paper on Bayesian model comparison. The function will return a value for 2\ln(B_{10}), where the model 1 is the first argument in the function and model 0 is the second. The table mentioned above is reproduced below.

    2\ln(B_{10}) B_{10} Interpretation
    0 to 2 1 to 3 Not worth a bare mention
    2 to 6 3 to 20 Positive
    6 to 10 20 to 150 Strong
    > 10 > 150 Very Strong

    So if uravu.utils.bayes_factor(model1, model2) returns 4.3, there is “Positive” evidence for model1 over model2.