So for now I can just use the mean value of the posterior distribution as a fixed value in the next model.