kcpm: Kernel of Imprecise Probability Measure Formulated By Bickis and Lee
Description
Imprecise probability density function proposed by Bickis and Lee (2014) is defined. See Details.
Usage
kcpm(t, xi2, xi1, xi0, log = FALSE)
kcpm_m(m, xi2, xi1, xi0, log = FALSE)
kcpm.ztrunc(m, xi2, xi1, xi0, ny, log = FALSE)
kcpm.ztrunc_t(t, xi2, xi1, xi0, ny, log = TRUE)
Arguments
t
random variable
xi2
parameter associated with precision
xi1
parameter associated with linear combination
xi0
parameter associated with effective sample size
log
logical; if TRUE (default), a returned value is given in logarithm scale.
m
random variable
ny
number of observations
Details
The formal definition of Bickis and Lee's conjugate formulation is
$$e^(-\xi_2\theta^2 + \xi_1\theta - \xi_0\exp(\theta))$$
$\theta$ is ranged from -Inf to Inf.
References
Lee, C.H. (2014) Imprecise Prior for Imprecise Inference on Poisson Sampling Model, PhD Thesis, Biostatistics Program, University of Saskatchewan