priorcontrol.dpm: Prior information for the 'pooledROC.dpm'

In ROCnReg: ROC Curve Inference with and without Covariates

Prior information for the pooledROC.dpm

Description

This function is used to set various parameters controlling the prior information to be used in the pooledROC.dpm function.

Usage

priorcontrol.dpm(m0 = NA, S0 = NA, a = 2, b = NA, alpha = 1, L = 10)

Arguments

m0	A numeric value. Hyperparameter; mean of the normal prior distribution for the means of each component. NA signals autoinitialization, with defaults: 0 if the data are standardised and \bar{y}_d (d \in \{D, \bar{D}\} if the data are not standardised.
S0	A numeric value. Hyperparameter; variance of the normal prior distribution for the means of each component. NA signals autoinitialization, with defaults: 10 if the data are standardised and 100*\frac{s^2_d}{n_d} (d \in \{D, \bar{D}\} if the data are not standardised, where s_d denotes the sample standard deviation.
a	A numeric value. Hyperparameter; shape parameter of the gamma prior distribution for the precisions (inverse variances) of each component. The default is 2.
b	A numeric value. Hyperparameter; rate parameter of the gamma prior distribution for the precisions (inverse variances) of each component. NA signals autoinitialization, with defaults: 0.5 if the data are standardised and \frac{s^2_d}{2} (d \in \{D, \bar{D}\} if the data are not standardised.
alpha	A numeric value. Precision parameter of the Dirichlet Process. The default is 1.
L	A numeric value. Upper bound on the number of mixture components. Setting L=1 corresponds to a normal model. The default is 10.

Value

A list with components for each of the possible arguments.

See Also

pooledROC.dpm

ROCnReg documentation built on March 31, 2023, 5:42 p.m.