Multidimensional Gauss-Hermite Quadrature

Uses a transformed, rotated and optionally adapted n-dimensional grid of quadrature points to calculate the numerical integral of n multivariate normal distributed parameters.


Multidimensional Gauss-Hermite quadrature in R.

The package consists of two functions, eval.quad() and init.quad(). The former evaluates the integral of any given function with respect to Q multivariate normal distributed parameters. The latter provides the quadrature points and weights at which the function should be evaluated, and the weights attached to each evaluation.

Quadrature points are taken from package fastGHQuad, and expanded into a multidimensional grid. This grid is then adjusted for a given multivariate normal distribution - usually a prior or population distribution. When this prior distribution is far from the posterior distribution under integration, precision of the estimate suffers. To overcome this, the quadrature grid can be further adjusted to adapt to an estimate of the posterior distribution.

© Karel Kroeze 2014.

Build Status

News

MultiGHQuad v1.2.0 (Release date: 2016-08-11)

  • CRITICAL: Fixed errors in eval.quad() pertaining to adaptation to intermediate results.
  • Added debug flag for init.quad(), provides a visual representation of the quadrature grid.

Reference manual

It appears you don't have a PDF plugin for this browser. You can click here to download the reference manual.

install.packages("MultiGHQuad")

1.2.0 by Karel Kroeze, 3 years ago


Browse source code at https://github.com/cran/MultiGHQuad


Authors: Karel Kroeze [aut, cre] , Rivka de Vries [ctb] , Alwin Stegeman [ctb]


Documentation:   PDF Manual  


GPL-3 license


Imports fastGHQuad

Depends on mvtnorm, Matrix

Suggests testthat


See at CRAN