conj.approx: Approximate distribution with mixture model

In StudyPrior: Distributions for Historical Information

Description

Approximate distribution with mixture model

Usage

conj.approx(distr, type = c("beta", "normal"), max.degree = 3,
  min.degree = max.degree, robust = FALSE, range = c(0, 1),
  do.plot = FALSE, return.value = 0.2, starts, length.fit = 100)

Arguments

distr	The density function of the function to be approximated
type	Use beta or normal densities in the mixture
max.degree	Maximum size of mixture
min.degree	Minimum size of mixture
robust	Include a robust component? (FALSE or final weight between 0 and 1)
range	The range of the density function to approximate
do.plot	Show the components of the mixture in a plot
return.value	Convergence threshold
starts	Initial parameters for the search
length.fit	Number of points to fit to

Value

Returns a mixture.prior object

StudyPrior documentation built on May 2, 2019, 5:54 p.m.