getStNeighbours: Creating Local Spatio-Temporal Neighbourhoods

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

Description

This function calculates local spatio-temporal neighbourhoods to be used for fitting of spatio-temporal vine copulas and for prediction using spatio-temporal vine copulas.

Usage

getStNeighbours(stData, ST, spSize = 4, tlags=-(0:2), var = names(stData@data)[1], 
                coVar=character(), timeSteps=NA, prediction=FALSE, min.dist = 0.01)

Arguments

stData	some spatio-temporal data frame holding the data used for estimation/prediction
ST	A spatio-temporal object defining the prediction locations, might be missing if the spatio-temporal neighbourhood is used for fitting.
spSize	The spatial size of the neighbourhood including the location of interest (for fitting as well for prediction).
tlags	The temporal lags to be used in the spatio-temporal neighbourhood.
var	the variable name of interest, by default the first variable is used
coVar	the covariate's name
timeSteps	The number of time instances that should randomly be selected from stData. The default, NA, selects all locations.
prediction	whether the neighbourhood should be used for prediction (TRUE) or spatial/Spatio-temporal vine copula fitting.
min.dist	the minimal distance for a location to be included. Must be larger than 0 for fitting purposes and might be 0 for prediction.

Value

An object of stNeighbourhood.

Author(s)

Benedikt Graeler

See Also

See stNeighbourhood for the native constructor of a stNeighbourhood class. The pure spatial version can be found at getNeighbours.

Examples

library("sp")
library("spacetime")

sp <- SpatialPoints(matrix(c(181000,181100,333500,333600),2))
time <- as.POSIXct("2014-03-18")+60*60*24*c(0,1,2)
data <- data.frame(measure=runif(6))

stData <- STFDF(sp, time, data)
stQuerry <- STF(SpatialPoints(matrix(c(181000,181200,333600,333600),2)),
                time[2:3])

getStNeighbours(stData=stData, ST=stQuerry, var="measure", spSize=3, 
                tlags=-(0:1), prediction=TRUE)

spcopula documentation built on May 2, 2019, 4:49 p.m.