demvn: Evaluate the density of a multivariate Gaussian fit
In esaddle: Extended Empirical Saddlepoint Density Approximations

Description Usage Arguments Details Value Author(s) Examples

View source: R/demvn.R

Given a sample X, it gives a pointwise evaluation of the multivariate normal (MVN) density fit at position y.

1	demvn(y, X, log = FALSE, verbose = TRUE, alpha = 2, beta = 1.25)

`y`	points at which the MVN is evaluated. It can be either a d-dimensional vector or an n by d matrix, each row indicating a different position.
`X`	an n by d matrix containing the data.
`log`	if TRUE the log-density is returned.
`verbose`	currently not used.
`alpha`	tuning parameter of `robCov`, see `?robCov` for details.
`beta`	tuning parameter of `robCov`, see `?robCov` for details.

The covariance matrix is estimated robustly, using the robCov function.

A vector where the i-th entry is the density corresponding to the i-th row of y.

Matteo Fasiolo <matteo.fasiolo@gmail.com> and Simon N. Wood.

1
2
3

library(esaddle)
X <- matrix(rnorm(2 * 1e3), 1e3, 2) # Sample used to fit a multivariate Gaussian
demvn(rnorm(2), X, log = TRUE)      # Evaluate the fitted log-density at a random location