I’ve read the Wald discussion, but I don’t fully understand how it relates to my problem. If i further simplify then this is what i am trying to build in pseudocode:
theta ~ Normal(mu=0,sd=1) # a prior on theta
q = f(y, theta) #a function that makes a nonlinear transformation of my observations (y) which depends on theta
q ~ Normal(mu=0,sd=1)
Does anybody know of any other examples where something similar is done?