matern.variog: Internal function used to estimate the nugget effect,...

In ProbForecastGOP: Probabilistic weather forecast using the GOP method

Description

Internal function called by the Variog.fit and the ProbForecastGOP functions when fitting a Whittle-Matern theoretical variogram to an empirical variogram via the method of Weighted Least Squares.

Usage

matern.variog(v,variog,d,w)

Arguments

v	numeric vector with the nugget effect, variance, range, and smoothness parameters of the Whittle-Matern variogram.
variog	numeric vector giving the values of the empirical variogram at distances given by the numeric vector d.
d	numeric vector giving the distances (or the bin midpoints) at which the empirical variogram has been computed.
w	numeric vector giving the weights to be used in the Weighted Least Squares, that is, the number of pairs of meteorological stations with distance falling in any given bin.

Details

This function is an internal function that is used and called by the Variog.fit and ProbForecastGOP functions to estimate the parameters of a Whittle-Matern variogram when fitting to an empirical variogram via Weighted Least Squares. - Defaults -

None.

Value

The function returns the weighted least-square loss function relative to the empirical variogram and the theoretical Whittle-Matern variogram evluated for a given set v of nugget effect, variance, range, and smoothness parameters. This is the function that is minimized to obtain estimates of the nugget effect, variance, range, and smoothness parameters of the Whittle-Matern variogram.

Author(s)

Berrocal, V. J. [e-mail], Gel, Y., Raftery, A. E., Gneiting, T.

ProbForecastGOP documentation built on May 2, 2019, 3:42 a.m.