t_to_mix: Generate an object of class 'normalmix' that is an...

Description Usage Arguments Author(s)

Description

This is based on the representation of the t as an inverse-gamma scaled-mixture of Gaussians.

Usage

1
t_to_mix(mu, sig, df, gridsize = 20, eps = 10^-6)

Arguments

mu

The mean of the t-distribution.

sig

The scale parameter of the t-distribution.

df

The degrees of freedom of the t-distribution.

gridsize

The number of mixture components to use. The larger the more accurate the approximation, but the higher the computational load — especially if you intend to use this for stramash.workhorse.

eps

A positive numeric. The tolerance from 0 and 1 for the grid in the probability space.

Author(s)

David Gerard


dcgerard/stramash documentation built on May 15, 2019, 1:24 a.m.