Fit a variogram model by Gaussian maximum likelihood

Fits (nugget, partial sill, range) of a spherical / exponential / Gaussian covariance model by maximizing the Gaussian log-likelihood of the (constant-mean) data, via stats::optim on a log parameterization.

Usage

morie_spatial_variogram_fit(coords, values, model = "exponential")

Arguments

coords, values

As in morie_spatial_variogram.

model

"exponential" (default), "spherical", or "gaussian".

Value

A list with model, nugget, psill, range, loglik, converged, method.

Examples

set.seed(1)
coords <- cbind(runif(60, 0, 10), runif(60, 0, 10))
values <- coords[, 1] * 0.5 + rnorm(60, 0, 0.3)
str(morie_spatial_variogram_fit(coords, values), max.level = 1)
#> List of 7
#>  $ model    : chr "exponential"
#>  $ nugget   : num 0.31
#>  $ psill    : num 2.47
#>  $ range    : num 33.9
#>  $ loglik   : num 5.22
#>  $ converged: logi TRUE
#>  $ method   : chr "variogram ML (rmorie native)"