Fit an exponential or gaussian variogram

Share:

Description

Fit an exponential or gaussian variogram with interdistances and corresponding dispersion values.

Usage

1
2
Fvariogfit3(disp.lt, h.lt, model = 1, a0 = 0.1, t0 = 0.5,
            max.iter = 25, max.fcal = 100, bep = 0, verbose = FALSE)

Arguments

disp.lt

vector of spatial dispersion values between locations (one value for each pair)

h.lt

vector of corresponding interdistances

model

type of variogram: 1 for exponential or 2 for gaussian

a0, t0

initial variogram parameter estimates

max.iter, max.fcal

control parameters for calls to non-linear optimization routines

bep

bending energy penalty, described in Sampson-Guttorp method (JASA 1992)

verbose

if TRUE, display fitting details to the console

Details

Exponential variogram is defined as a_1 + (2-a_1) (1-exp(-t_0 h)). Gaussian variogram is defined as a_1 + (2-a_1) (1-exp(-t_0 h^2)).

Value

A list containing the following named components:

objf

sum of residual sum of squares, plus bending energy penalty

a, t0

fitted parameters

fit

fitted values of variogram