At the moment, if one performs the bayesian update, the second fit does not return the total evidence but only the second piece of the integral. The total evidence could be recovered by summing the two log evidences, but the code needs to be enhanced for this. We could consider having some automatic handling of it, rather than leaving it to the user.
At the moment, if one performs the bayesian update, the second fit does not return the total evidence but only the second piece of the integral. The total evidence could be recovered by summing the two log evidences, but the code needs to be enhanced for this. We could consider having some automatic handling of it, rather than leaving it to the user.