sim.weed {WeedMap} | R Documentation |
The data simulated are of two kinds: exact weed counts and image derived indexes. The counts w arise from a transformed Gaussian model with Gamma marginal. The image indexes i relate to w as i(s) = λ w(s) * ε(s), where ε is a spatially uncorrelated noise. The simulation is carried out on the unit square.
sim.weed(nx, ny, nxy, nz, param.cov, mu, sigma, lambda, tau, nbin, true.field = FALSE, npix = NULL, z.on.grid = TRUE)
nx |
number of sites where only w is observed |
ny |
number of sites where only i is observed |
nxy |
number of sites where w and i are observed |
nz |
number of sites where w will be predicted. If
z.on.grid=TRUE , nz should be a square. |
param.cov |
Vector of parameters of the underlying Gaussian
randomFields. The simulation of the Gaussian random fields performed
assuming an exponential covariance function. The vector of
parameters should have four components, namely: mea, variance,
nugget, scale and should be given as
e.g.:c(mean=0,variance=1,nugget=0,scale=.1) . See the
documentation of GaussRF in package
RandomFields for details. |
mu |
Mean of w |
sigma |
Variance of w |
lambda |
Scaling factor relating w and i |
tau |
Variance of ε |
nbin |
Number of bins for the binned data |
true.field |
Logical: shoud values on a dense grid be given |
npix |
A vector giving the number of pixels horizontally and
vertically of the grid if true.field=TRUE |
z.on.grid |
Logical: set TRUE if sites of prediction z are
required on a grid; thennz should be a square. |
x |
Coordinates of sites where weed counts only are given |
xy |
Coordinates of sites where weeds and images are given |
y |
Coordinates of sites where image indexes only are given |
z |
Coordinates of sites where predcited values are sought |
coord.grid |
Logical telling whether all the realisation of the weed random field is also given on a grid |
wx |
Weed values at sites x |
wy |
Weed values at sites y |
wxy |
Weed values at sites xy |
wz |
Weed values at sites z |
vx |
Binned weed values at sites x |
vxy |
Binned weed values at sites xy |
i |
Image values at sites xy and y . Values are
concatenated in the same vector in this order |
wgrid |
Weed counts values at the node of a grid |
igrid |
Image index values at the node of a grid |
bin |
Binning of the weed counts |
param.cov |
Parameters of the covariance function of the underlying Gaussian random field |
mu |
Mean of w |
sigma |
Variance of w |
alpha |
Shape parameter of w |
beta |
Scale parameter of w |
lambda |
Scaling parameter relating w and i as
i = λ w ε |
tau |
Variance of noise epsilon |
npix |
number of pixel in the horizontal direction (same value is assumed for the vertical direction) |
Gilles Guillot
G. Guillot, N. Loren, M. Rudemo, Bayesian spatial prediction of weed intensities from exact count data and picture based indexes, 2006, submitted