Description Usage Arguments Details Value Author(s) Examples
mgp
handles the evaluation of multivariate Gaussian processes with two
behaviours. It evaluates the multivariate effects applied to the coordinates
coords
if size == NULL
or it simulates the coordinates
coords
if size
is provided.
1 2 |
coords |
A list of replicated coordinates. Each coordinate is a q replicated
vector of the unique coordinates. If |
A |
A qxm matrix that defines the relationship between Y(h) and S(h). |
cor.model |
A character or function indicating the correlation function to be used to compute the correlation matrix |
cor.params |
A nested list of the parameters of the correlation function for each response. |
size |
A numeric value n representing the number of locations, it is used to
simulate the coordinates |
range |
Range of the coordinates of the Gaussian processes. |
geom |
object of class ‘sf’ or ‘sfc’ to define the area where to simulate the samples |
The q-dimensional multivariate Gaussian process Y(h) is expressed as a linear combination
Y(h) = AS(h).
where S(h) is a vector of m standardized independent Gaussian processes at location h and A a qxm matrix of coefficients.
A list of simulated replicated coordinates in case size
is provided;
otherwise, a nq-length numeric vector of the evaluated multivariate Gaussian
process.
Erick A. Chacón-Montalván
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 | ## Simulation of a temporal multivariate Gaussian process
# Define the number of location and responses, and a linear transformation matrix
n <- 200
q <- 3
A <- matrix(c(1, 0, 1, 0, 1, - 0.8), nrow = q)
# Simulate coordinates
coords <- mgp(list(time = NA), A, size = n)
# Simulate a multivariate Gaussian process
cor.params <- list(list(phi = 0.08), list(phi = 0.15))
y <- mgp(coords, A, "exp_cor", cor.params)
# Visualize the temporal multivariate Gaussian process
data <- data.frame(time = coords[[1]], y, response = factor(rep(1:q, each = n)))
library(ggplot2)
ggplot(data, aes(time)) +
geom_line(aes(y = y, col = response))
## Simulation of a spatial multivariate Gaussian process
# Define the number of location and responses, and a linear transformation matrix
n <- 1500
q <- 2
A <- matrix(c(1, -1), nrow = q)
# Simulate coordinates
coords <- mgp(list(s1 = NA, s2 = NA), A, size = n)
# Simulate a multivariate Gaussian process
cor.params <- list(list(phi = 0.1))
y <- mgp(coords, A, "exp_cor", cor.params)
# Visualize the temporal multivariate Gaussian process
data <- data.frame(s1 = coords[[1]], s2 = coords[[2]], y,
response = factor(rep(1:q, each = n)))
ggplot(data, aes(s1, s2)) +
geom_point(aes(size = y, col = y)) +
facet_wrap(~ response)
|
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.