condSpVine: Conditions a spatial vine copula for conditional prediction

In BenGraeler/spcopula: Copula Driven Analysis - Multivariate, Spatial, Spatio-Temporal

Description

A spatial vine copula is conditioned under the observations of all but one neighbour generating a conditional univariate distribution used for prediction.

Usage

condSpVine(condVar, dists, spVine, n = 1000)

Arguments

condVar	the conditional variables
dists	the distances between the neighbours to adjust the spatial vine copula parameters.
spVine	the spatial vine copula
n	a proxy to the number of intervals to be used in the numerical integration. The intervals will not be split uniform with a shorter intervals at the corners of the copula.

Value

A function describing the conditional density.

Author(s)

Benedikt Graeler

See Also

spVineCopula

Examples

data("spCopDemo")

calcKTauPol <- fitCorFun(bins, degree=3)

spCop <- spCopula(components=list(normalCopula(0.2), tCopula(0.2, dispstr = "un"),
                                  frankCopula(1.2), normalCopula(0.2), claytonCopula(0.2),
                                  claytonCopula(0.2), claytonCopula(0.2), claytonCopula(0.2),
                                  claytonCopula(0.2), indepCopula()),
                  distances=c(0, bins$meanDists[1:9]),
                  spDepFun=calcKTauPol, unit="m")

spVineCop <- spVineCopula(spCop, vineCopula(4L))

dists <- list(c(473, 124, 116, 649))
condVar <- c(0.29, 0.55, 0.05, 0.41)
condDensity <- condSpVine(condVar,dists,spVineCop)

curve(condDensity)
mtext(paste("Dists:",paste(round(dists[[1]],0),collapse=", ")),line=0)
mtext(paste("Cond.:",paste(round(condVar,2),collapse=", ")),line=1)

BenGraeler/spcopula documentation built on Nov. 20, 2020, 4:07 p.m.