Hi,
I was going through some examples under the “change of variables” topic. Compiling the models from those examples should, in principle, lead to warnings such as:
DIAGNOSTIC(S) FROM PARSER:Warning (non-fatal):
Left-hand side of sampling statement (~) may contain a non-linear transform of a parameter or local variable.
...
The problem is that I don’t get these warnings. Otherwise the models compile and run and the results are as expected.
For example, I tried with the following model from the paper on The Impact of Reparameterization on Point Estimates by Bob Carpenter.
data {
int<lower=0> N;
int<lower=0, upper=1> y[N];
}
parameters {
real alpha;
}
transformed parameters {
real<lower=0, upper=1> theta;
theta <- inv_logit(alpha);
}
model {
for (n in 1:N)
y[n] ~ bernoulli(theta);
theta ~ uniform(0, 1);
}
The results from the paper are replicated, but with no warning about the need for Jacobian adjustment.
I use R 4.2.1 with rstan_2.26.16. I do get to see other (unrelated) Stan warnings, so they are not totally suppressed.
Will be grateful for any suggestions.
Jakub