bayesian - Hyperpriors for hierarchical models with Stan -
i'm looking fit model estimate multiple probabilities binomial data stan. using beta priors each probability, i've been reading using hyperpriors pool information , encourage shrinkage on estimates.
i've seen example define hyperprior in pymc, i'm not sure how similar stan
@pymc.stochastic(dtype=np.float64) def beta_priors(value=[1.0, 1.0]): a, b = value if <= 0 or b <= 0: return -np.inf else: return np.log(np.power((a + b), -2.5)) = beta_priors[0] b = beta_priors[1]
with , b being used parameters beta prior.
can give me pointers on how similar done stan?
following suggestions in comments i'm not sure follow approach, reference thought i'd @ least post answer question of how accomplished in stan.
after asking around on stan discourses , further investigation found solution set custom density distribution , use target +=
syntax. equivalent stan of example pymc be:
parameters { real<lower=0> a; real<lower=0> b; real<lower=0,upper=1> p; ... } model { target += log((a + b)^-2.5); p ~ beta(a,b) ... }
Comments
Post a Comment