Thanks for the tip. I will read the Stan manual to see if I can follow the examples about using censoring for my problem.
I have a more general question about modeling processes of the following type. Assume
z1 ~ Normal(mu1, sigma1)
z2 ~ Normal(mu2, sigma2)
x = f(z1, z2)
where f
is a deterministic function.
If I observe x
and am trying to infer E[z1, z2 | x]
, how do I go about modeling this in Stan?
Looks like I’ll have to put x
in the data section, but when I try doing that but declare x = f(z1, z2)
in the model section I get the “cannot assign variable outside declaration block” error.