switch 8 schools to non-centered parameterization?

[bob-carpenter]

Either 8 schools should be swapped out for something simpler or easy to fit, or it should be changed to non-centered parameterization. Or if it's not meant to fit, that should be somehow made apparent to users. (If it is, feel free to summarily close this issue.)

From Vishv Jeet on stan-users:

> library(rstan)
...
rstan (Version 2.14.1, packaged: 2016-12-28 14:55:41 UTC, GitRev: 5fa1e80eb817)
...
1: package ‘rstan’ was built under R version 3.2.5 
2: package ‘ggplot2’ was built under R version 3.2.5 
3: package ‘StanHeaders’ was built under R version 3.2.5 

> fit <- stan_demo('eight_schools')

Loading required namespace: rstudioapi
Warning messages:
1: There were 145 divergent transitions after warmup. Increasing adapt_delta above 0.8 may help. See
http://mc-stan.org/misc/warnings.html#divergent-transitions-after-warmup 
2: There were 1 chains where the estimated Bayesian Fraction of Missing Information was low. See
http://mc-stan.org/misc/warnings.html#bfmi-low 
3: Examine the pairs() plot to diagnose sampling problems

When looked thru the shinystan I found the model code was non centered indeed.

data {
  int<lower=0> J;          // number of schools
  real y[J];               // estimated treatment effect (school j)
  real<lower=0> sigma[J];  // std err of effect estimate (school j)
}
parameters {
  real mu;
  real theta[J];
  real<lower=0> tau;
}
model {
  theta ~ normal(mu, tau); 
  y ~ normal(theta,sigma);
}

[bgoodri]

It should be fixed in example-models.

[jgabry]

It would probably also be good to have both parameterizations as examples.

[bob-carpenter]

Especially for ShinyStan to highlight divergences! - Bob

…

On Feb 6, 2017, at 3:21 PM, Jonah Gabry ***@***.***> wrote: It would probably also be good to have both parameterizations as examples. — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub, or mute the thread.