dmvnorm_cholesky_inverse: Evaluate density of multivariate Normal
In pierrejacob/PET:

Description Usage Arguments Details Value See Also Examples

View source: R/multivariatenormal.R

Function to evaluate a multivariate Normal density of each row of the given matrix, where the covariance matrix is specified through a Cholesky factor of its inverse.

1	dmvnorm_cholesky_inverse(x, mean, cholesky_inverse)

`x`	is a n x d matrix of real values
`mean`	is the mean vector
`cholesky_inverse`	is a Cholesky factor of the inverse of the covariance matrix; for instance obtained as `t(chol(solve(covariance)))`.

This function does not check anything (i.e. that the given covariance is PSD). Thus it is faster than functions in standard packages, but more risky.

a vector of n log-density evaluations

rmvnorm, rmvnorm_cholesky, dmvnorm

1
2
3

x <- rmvnorm(10, rep(0, 5), diag(1, 5, 5))
dmvnorm_cholesky_inverse(x, rep(-1, 5), t(chol(solve(diag(2, 5, 5)))))
# mvtnorm::dmvnorm(x, rep(-1, 5), diag(2, 5, 5), log = T)