default_prior is a generic function that can be used to
get default priors for Bayesian models. Its original use is
within the brms package, but new methods for use
with objects from other packages can be registered to the same generic.
default_prior(object, ...)
get_prior(formula, ...)Usually, a brmsprior object. See
default_prior.default for more details.
See default_prior.default for the default method applied for
brms models. You can view the available methods by typing
methods(default_prior).
## get all parameters and parameters classes to define priors on
(prior <- default_prior(count ~ zAge + zBase * Trt + (1|patient) + (1|obs),
data = epilepsy, family = poisson()))
#> prior class coef group resp dpar nlpar lb ub tag
#> student_t(3, 1.4, 2.5) Intercept
#> (flat) b
#> (flat) b Trt1
#> (flat) b zAge
#> (flat) b zBase
#> (flat) b zBase:Trt1
#> student_t(3, 0, 2.5) sd 0
#> student_t(3, 0, 2.5) sd obs 0
#> student_t(3, 0, 2.5) sd Intercept obs 0
#> student_t(3, 0, 2.5) sd patient 0
#> student_t(3, 0, 2.5) sd Intercept patient 0
#> source
#> default
#> default
#> (vectorized)
#> (vectorized)
#> (vectorized)
#> (vectorized)
#> default
#> (vectorized)
#> (vectorized)
#> (vectorized)
#> (vectorized)