Create an object representing an inverse Wishart prior, possibly with modeled scale matrix
pr_invwishart(df = NULL, scale = NULL)
An environment representing the specified prior, for internal use.
Degrees of freedom parameter. This should be a scalar numeric value. Default value is the dimension plus one.
Either a (known) scale matrix, or
scale="modeled"
or scale="modelled"
, which puts default
chi-squared priors on the diagonal elements of the inverse Wishart scale matrix.
For more control on these chi-squared priors a list can be passed with some of the
following components:
degrees of freedom (scalar or vector) of the chi-squared distribution(s)
scale parameter(s) of the chi-squared distribution(s)
whether the modeled scale parameter of the inverse chi-squared
distribution is (a scalar parameter) common to all n
diagonal elements.
A. Huang and M.P. Wand (2013). Simple marginally noninformative prior distributions for covariance matrices. Bayesian Analysis 8, 439-452.