eval_basis function

Evaluate basis functions

Evaluate basis functions at points or average functions over polygons.

eval_basis(basis, s)

## S4 method for signature 'Basis,matrix'
eval_basis(basis, s)

## S4 method for signature 'Basis,SpatialPointsDataFrame'
eval_basis(basis, s)

## S4 method for signature 'Basis,SpatialPolygonsDataFrame'
eval_basis(basis, s)

## S4 method for signature 'Basis,STIDF'
eval_basis(basis, s)

## S4 method for signature 'TensorP_Basis,matrix'
eval_basis(basis, s)

## S4 method for signature 'TensorP_Basis,STIDF'
eval_basis(basis, s)

## S4 method for signature 'TensorP_Basis,STFDF'
eval_basis(basis, s)

Arguments

• basis: object of class Basis
• s: object of class matrix, SpatialPointsDataFrame or SpatialPolygonsDataFrame containing the spatial locations/footprints

Details

This function evaluates the basis functions at isolated points, or averages the basis functions over polygons, for computing the matrix $S$. The latter operation is carried out using Monte Carlo integration with 1000 samples per polygon. When using space-time basis functions, the object must contain a field t containing a numeric representation of the time, for example, containing the number of seconds, hours, or days since the first data point.

Examples

library(sp)

### Create a synthetic dataset
set.seed(1)
d <- data.frame(lon = runif(n=500,min = -179, max = 179),
                lat = runif(n=500,min = -90, max = 90),
                z = rnorm(500))
coordinates(d) <- ~lon + lat
slot(d, "proj4string") = CRS("+proj=longlat")

### Now create basis functions on sphere
G <- auto_basis(manifold = sphere(),data=d,
                nres = 2,prune=15,
                type = "bisquare",
                subsamp = 20000)

### Now evaluate basis functions at origin
S <- eval_basis(G,matrix(c(0,0),1,2))

See Also

auto_basis for automatically constructing basis functions.