Create an object of class 'HyperparametersBatch' for the batch mixture model

Share:

Description

Create an object of class 'HyperparametersBatch' for the batch mixture model

Usage

1
2
HyperparametersBatch(k = 0L, mu.0 = 0, tau2.0 = 100, eta.0 = 1800,
  m2.0 = 1/60, alpha, beta = 0.1, a = 1.8, b = 6)

Arguments

k

length-one integer vector specifying number of components (typically 1 <= k <= 4)

mu.0

length-one numeric vector of the of the normal prior for the component means.

tau2.0

length-one numeric vector of the variance for the normal prior of the component means

eta.0

length-one numeric vector of the shape parameter for the Inverse Gamma prior of the component variances, tau2_h. The shape parameter is parameterized as 1/2 * eta.0. In the batch model, tau2_h describes the inter-batch heterogeneity of means for component h.

m2.0

length-one numeric vector of the rate parameter for the Inverse Gamma prior of the component variances, tau2_h. The rate parameter is parameterized as 1/2 * eta.0 * m2.0. In the batch model, tau2_h describes the inter-batch heterogeneity of means for component h.

alpha

length-k numeric vector of the shape parameters for the dirichlet prior on the mixture probabilities

beta

length-one numeric vector for the parameter of the geometric prior for nu.0 (nu.0 is the shape parameter of the Inverse Gamma sampling distribution for the component-specific variances. Together, nu.0 and sigma2.0 model inter-component heterogeneity in variances.). beta is a probability and must be in the interval [0,1].

a

length-one numeric vector of the shape parameter for the Gamma prior used for sigma2.0 (sigma2.0 is the shape parameter of the Inverse Gamma sampling distribution for the component-specific variances).

b

a length-one numeric vector of the rate parameter for the Gamma prior used for sigma2.0 (sigma2.0 is the rate parameter of the Inverse Gamma sampling distribution for the component-specific variances)

Value

An object of class HyperparametersBatch

Examples

1

Want to suggest features or report bugs for rdrr.io? Use the GitHub issue tracker.