VeccTMVN (version 1.3.1)

ptmvrandn: Simulate partially censored multivariate normal (MVN) at censored locations using the Vecchia approximation

Description

Simulate partially censored multivariate normal (MVN) at censored locations using the Vecchia approximation

Usage

ptmvrandn(
  locs,
  indCensor,
  y,
  bCensor,
  covName = NULL,
  covParms = NULL,
  m = 30,
  N = 1000,
  verbose = TRUE,
  reorder = TRUE
)

Value

n X N matrix of generated samples

Arguments

locs

location (feature) matrix n X d

indCensor

indices of locations that have only censored observations

y

observed (not censored) values, of length n

bCensor

upper bound, above which observations are not censored, can be different for different locations, of length 1 or n

covName

covariance function name from the `GpGp` package

covParms

parameters for `covName`

m

Vecchia conditioning set size

N

number of samples required

verbose

verbose level

reorder

whether to Vecchia univariate variable reordering