These functions are deprecated and will be removed in a future release. The Arguments section below provides details on how the functionality obtained via each of the arguments has been replaced.

prior_options(
  prior_scale_for_dispersion = 5,
  min_prior_scale = 1e-12,
  scaled = TRUE
)

Arguments

prior_scale_for_dispersion, min_prior_scale, scaled

Arguments to deprecated prior_options function. The functionality provided by the now deprecated prior_options function has been replaced as follows:

prior_scale_for_dispersion

Instead of using the prior_scale_for_dispersion argument to prior_options, priors for these parameters can now be specified directly when calling stan_glm (or stan_glmer, etc.) using the new prior_aux argument.

scaled

Instead of setting prior_options(scaled=FALSE), internal rescaling is now toggled using the new autoscale arguments to normal, student_t, and cauchy (the other prior distributions do not support 'autoscale').

min_prior_scale

No replacement. min_prior_scale (the minimum possible scale parameter value that be used for priors) is now fixed to 1e-12.